The status field contains either CLOSED, FIXED or OPEN.
These pages were automatically created by a perl script called
make_issues.pl that is freely available by clicking
here the input to the script is a text file.
PAGE_DESC_END:
# ================================================
ENTRY_ID: 0001
ENTRY_TITLE: The -rootfile switch parses relative names improperly.
ENTRY_REPORTED_BY: Dékány Dániel
ENTRY_REPORTED_ON: 2001/08/22
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2001/08/22
ENTRY_REPORTED_BEGIN:
Little bug: Switch '-rootfile': If there is no parh given, the first
letter of filename will missing from the HTML hrefs. Eg:
'-rootfile index.html' => '<a href='ndex.html'>@root</a>' ('i' missing)
but
'-rootfile ./index.html' => '<a href='index.html'>@root</a>'
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Found a problem in make_file_url. The separator iterator was being
incremented when the path was completely relative.
// ================================================================
// Make file URL, strip off the leading directories.
// ================================================================
void ccdoc::phase3::html::make_file_url(string& url,string& fn) const
{
// Strip off the leading directories to
// make the relative reference work.
string::iterator itr = fn.begin();
string::iterator relpath_itr = fn.begin();
for(;itr!=fn.end();++itr) {
if( '/' == *itr ) {
relpath_itr = itr;
}
}
url = "";
for(++relpath_itr;relpath_itr!=fn.end();++relpath_itr) {
url += *relpath_itr;
}
}
// ================================================================
// Make file URL, strip off the leading dSeptember 18, 2001 (Tuesday)
The new release of ccdoc is available: "ccdoc v0.8 r18 2001/09/18".
This release fixes issues 0052 and is the first release that contains the linux port.
irectories.
// ================================================================
void ccdoc::phase3::html::make_file_url(string& url,string& fn) const
{
// Strip off the leading directories to
// make the relative reference work.
string::iterator itr = fn.begin();
string::iterator relpath_itr = fn.begin();
for(;itr!=fn.end();++itr) {
if( '/' == *itr ) {
relpath_itr = itr;
++relpath_itr; // ISSUE 001 FIX
}
}
url = "";
for(;relpath_itr!=fn.end();++relpath_itr) {
url += *relpath_itr;
}
}
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0002
ENTRY_TITLE: The ccdoc_top tage is incorrectly placed in the HEAD section.
ENTRY_REPORTED_BY: Dékány Dániel
ENTRY_REPORTED_ON: 2001/08/22
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2001/08/22
ENTRY_REPORTED_BEGIN:
Very little bug: <a name=ccdoc_top></a> is placed into the HEAD, which is
not correct (according to HTML 4.0 at least, but I don't think its right for
HTML 3.x). It should be pleced just *before* that <table border=0
width="100%"> in the BODY. However I don't know a browser which is confused
by this.
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Moved <a name=ccdoc_top></a> to the BODY section.
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0003
ENTRY_TITLE: Need a new switch, similar to -head, that inserts a file after the head tag to insert custom META elements.
ENTRY_REPORTED_BY: Dékány Dániel
ENTRY_REPORTED_ON: 2001/08/22
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2001/08/22
ENTRY_REPORTED_BEGIN:
What is missing, a switch similar to '-head' which inserts the
specified file just after the <head> tag. This is required to insert custom
META elements, primary for set the charset. (Note that most browsers accepts
META http-equivs in the BODY (!) so I can use -head, but its not correct
HTML.)
(btw, if some users will report CRC errors in the gzipped files, this is because there is some conflict between the HTTP server and some (or only one?) proxy-es. Dont ask why, I got partial-content (HTTP 206) without the last few kbytes of files (=> gunzip detects CRC error), then decided to bypass the cache-proxy.)
C:> ccdoc.exe -db LB.db -pkg LBTools *h
WARNING: File '*.h' cannot be read so it will be ignored.
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r10.
I modified the MSVC version of the program to attempt to process wildcards if the attempt to open the specified file fails. The -files switch still works.
When I originally closed this issue, I used the following reasoning.
ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0008 ENTRY_TITLE: Turn off the ccdoc generation of special members. ENTRY_REPORTED_BY: Martin Trappel ENTRY_REPORTED_ON: 2001/08/23 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/23 ENTRY_REPORTED_BEGIN: Ccdoc seems to generate automatic comments for Standard Copy Constructor, Destructor and Assignment Operator.Wildcards do not work in the cmd.exe shell on Windows platforms because it does not expand them. Instead it expects the application to do the expansion.
Ccdoc cannot do that because a user can run many different shells on Windows platforms, such as: ksh, rsh and tcsh (I use tcsh). Most of those shells (in fact all of them execept cmd.exe) expand wildcards before passing the the program.
Unfortunately, ccdoc can't tell the difference a legitimate file name with a * or ? in it from the tcsh and a wildcard request from the cmd sheel so you have to tell it using the -files switch.
You can work around this problem by using the new -files switch as follows:
C:> DIR \B *.h > hfiles.txt C:> ccdoc -db foo.db -pkg foo -files hfiles.txt
Can I disable this, since I have a few classes where I do not have these and I don't quite see the point in having doc for non-existant members?
I agree with your point that you should not be forced to see them unless you want to and will look into providing a switch that gives you the control to turn the reporting of them on or off. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Added the -[no]cdsm switches to phase 1 to allow users to control whether or not default special members are created. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0009 ENTRY_TITLE: Compile problem in gcc 3.0 on SuSE linux platform. ENTRY_REPORTED_BY: Michael Davis ENTRY_REPORTED_ON: 2001/08/23 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/23 ENTRY_REPORTED_BEGIN: We use gcc 3.0 on a SuSE linux platform here, and your code compiled with only one glitch. Here is a patch should you want it:
*** phase3_html.cc.orig Thu Aug 23 09:52:22 2001
--- phase3_html.cc Thu Aug 23 09:36:43 2001
*************** void ccdoc::phase3::html::write_class_de
*** 476,482 ****
--xitr;
make_tag_id(*xitr,prev_tag);
}
! if( itr ) {
vector<statement::base*>::iterator xitr = itr;
++xitr;
if( xitr != contents.end() ) {
--- 476,482 ----
--xitr;
make_tag_id(*xitr,prev_tag);
}
! if( itr != contents.end() ) {
vector<statement::base*>::iterator xitr = itr;
++xitr;
if( xitr != contents.end() ) {
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Thank you. I have updated and released the fix.
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0010
ENTRY_TITLE: Random character appears between package name and function/macro/etc. name in the HTML title.
ENTRY_REPORTED_BY: Dékány Dániel
ENTRY_REPORTED_ON: 2001/08/23
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2001/08/23
ENTRY_REPORTED_BEGIN:
Random character appears between package name and function/macro/etc. name
in the HTML title.
Eg. this is the title for function RcMalloc in rcheap.h:
<title>ccdoc RcMalloc rcheap.h:C:function</title>'C' should not be there. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: This problem occurred because the lineno was being appended to the output string used for the title as an integer rather than a string in phase3_html.cc at line 2169. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0011 ENTRY_TITLE: Instantiations of class template methods are reported as global functions. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/08/24 ENTRY_STATUS: CLOSED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/24 ENTRY_REPORTED_BEGIN: In the following example, the instantiation of the template constructor is incorrectly reported as a global function.
template <class T>
class A {
public:
A();
};
// Instantiation is incorrectly reported as
// a global function.
template <class T>
A<T>::A()
{
}
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
This occurs because ccdoc attempts to do a name resolution to determine
where the instantiation belongs so it tries to look up A<T> and can't
find it.
The best way to work around this problem is to tell ccdoc to ignore the template method instantiations as shown below:
template <class T>
class A {
public:
A();
};
#if !defined(__ccdoc__)
// Instantiation is incorrectly reported as
// a global function.
template <class T>
A<T>::A()
{
}
#endif
Meanwhile, I am looking at adding a feature to ccdoc to tell it to
ignore template method instantiations (without ignoring template
functions), see issue 0012.
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0012
ENTRY_TITLE: Ccdoc should ignore template method instantiations.
ENTRY_REPORTED_BY: Joe Linoff
ENTRY_REPORTED_ON: 2001/08/24
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2003/02/17
ENTRY_REPORTED_BEGIN:
See issue 0011 for details of the problem.
Ccdoc should have a switch that allows users to tell it to turn
off (or on) the recognition of template method instantiations.
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r33 and verified by test 69.
See issue 0149 for more details. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0013 ENTRY_TITLE: Add ccdoc to source forge. ENTRY_REPORTED_BY: Scott MacDiarmid ENTRY_REPORTED_ON: 2001/08/24 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2002/01/02 ENTRY_REPORTED_BEGIN: I am sure you have been asked this before, but have you ever considered moving this project to SourceForge (http://www.sf.net)? It provides bug tracking, forums, and version control. I find it useful from both sides of the coin. Meaning as a developer and an end-user. As a developer it allows people to easy submit bugs to me, ask me question and it also provides CVS for good version control. As an end user I can follow what a project is doing very easily. The only catch is you would have to put it under an open source license. Anyway, just a thought.
/** Zero-argument const method type for Node */
typedef Node* (Node:*NodeConstMeth) () const;
Ccdoc will generate an entry with 'const' instead of NodeConstMeth.
void PlotDown(const doc::CC* pCC) const
throw (RunTimeError);
void PlotDown(const doc::CC* pCC) const
#if !defined(__ccdoc__)
throw (RunTimeError)
#endif
;
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r5.
This problem occurred because the get_fct_id() heuristic in phase1_parser assumed that the trailing set of outermost parentheses defined the function argument list. I changed it look for the trailing set of outermost parentheses before a "throw" keyword. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0016 ENTRY_TITLE: Source file names of the form ../folder/file.ext don't work properly in the cmd.exe shell under windows. ENTRY_REPORTED_BY: Yuriy Stul ENTRY_REPORTED_ON: 2001/08/26 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/28 ENTRY_REPORTED_BEGIN: Source file names of the form ../folder/file.exe don't work properly. The error message that it report is:
The instruction at "0x00402330" referenced memory at "0x6e64658f". The memory could not be read.
Here are two examples of the command lines I was testing:
..\bin_opt_msvc_MSWin32-4.0\ccdoc -db ..\test\test.db -html ..\test\html\ ..\sources\s1.h ..\sources\s2.h AND ..\bin_opt_msvc_MSWin32-4.0\ccdoc -db ../test/test.db -html ../test/html/ ../sources/s1.h ../sources/s2.hI tried multiple levels of directories. I tried different source file extensions and everything else I could think of.
All of this testing was done under Win2K using the MS-DOS command interpreter (cmd.exe).
Can you help figure out what am I doing wrong so that I can track down and fix this problem? ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r8.
First, my thanks to Yuriy for helping me track this down.
This was a different manifestation of issue 0020. The fix was the same. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0017 ENTRY_TITLE: Can't differentiate between different revisions of ccdoc 0.8. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/08/26 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/26 ENTRY_REPORTED_BEGIN: It is difficult to distinguish between different versions of ccdoc 0.8.
Added the -version switch. Deprecated the -cid switch. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0018 ENTRY_TITLE: Directory separators in -root names cause run time errors. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/08/27 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/27 ENTRY_REPORTED_BEGIN: Directory separators in -root names cause run time errors. If you use a directory separator in a -root name, it confuses ccdoc into thinking there is a directory when there isn't which causes a run-time error. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r6.
Added '/' and '\' filtering in ccdoc::phase3::html::format_name(). ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0019 ENTRY_TITLE: The @see directive no longer accepts user defined links. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/08/27 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/27 ENTRY_REPORTED_BEGIN: The @see directive no longer accepts user defined links. In the previous version of ccdoc I could put in a direct link as follows:
@see <a href=mywebpage.html>Useful Stuff</a>
Modified the scanner to accept @see arguments that begin with a '<'. Modified the html generator to output these without interpretation. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0020 ENTRY_TITLE: ccdoc crashes ENTRY_REPORTED_BY: Chi-Fung Fan ENTRY_REPORTED_ON: 2001/08/27 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/28 ENTRY_REPORTED_BEGIN: Both MSVC and cygwin versions of ccdoc crashed on generating docs for my program source. Attached is the debug messages before ccdoc dumped core.
I noticed that the offendng file that caused the core dump is the last file 'src/ScribDoc.h'. If I move it to be the first input file, everythinkg runs ok. This is also the workaround for me.
What information do you need to debug this problem?
ccdoc -v -db doc/webdocs/ccdoc.db -pkg Util include/mathpack/convolution.h include/mathpack/geometry.h include/ctl/iterator.h src/plot.h
db: read begins
db: does not exist
db: read ends
phase1: begins
phase1: parsing 'include/mathpack/convolution.h' ...
phase1: parsing 'include/mathpack/geometry.h' ...
phase1: parsing 'include/ctl/iterator.h' ...
phase1: parsing 'src/plot.h' ...
phase1: ends
db: write begins
db: writing 146 statements
db: write ends
ccdoc -v -db doc/webdocs/ccdoc.db -pkg Scribble src/navigator.h src/pager.h src/charannotvw.h src/drawtool.h src/presenter.h
db: read begins
db: read ends
phase1: begins
phase1: parsing 'src/navigator.h' ...
phase1: parsing 'src/pager.h' ...
phase1: parsing 'src/charannotvw.h' ...
phase1: parsing 'src/drawtool.h' ...
phase1: parsing 'src/presenter.h' ...
phase1: ends
db: write begins
db: writing 406 statements
db: write ends
ccdoc -v -db doc/webdocs/ccdoc.db -pkg Scribble src/definitions.h src/annottempl.h src/annotation.h src/chardefinitions.h src/dynamics.h
src/doctypes.h src/visitor.h src/docreader.h src/docwriter.h src/tablet.h src/glyph.h src/cc.h src/token.h src/page.h src/ScribDoc.h
db: read begins
db: read ends
phase1: begins
phase1: parsing 'src/definitions.h' ...
phase1: parsing 'src/annottempl.h' ...
phase1: parsing 'src/annotation.h' ...
phase1: parsing 'src/chardefinitions.h' ...
phase1: parsing 'src/dynamics.h' ...
phase1: parsing 'src/doctypes.h' ...
phase1: parsing 'src/visitor.h' ...
phase1: parsing 'src/docreader.h' ...
phase1: parsing 'src/docwriter.h' ...
phase1: parsing 'src/tablet.h' ...
phase1: parsing 'src/glyph.h' ...
phase1: parsing 'src/cc.h' ...
phase1: parsing 'src/token.h' ...
phase1: parsing 'src/page.h' ...
phase1: parsing 'src/ScribDoc.h' ...
phase1: ends
db: write begins
db: writing 1994 statements
db: write ends
0 [main] ccdoc 94223783 open_stackdumpfile: Dumping stack trace to CCDOC.EXE.stackdump
/cygdrive/c/CHIFUNG/BIN/mkscribdocs.sh: line 63: 94223783 Segmentation fault (core dumped) ccdoc -db doc/webdocs/ccdoc.db
root.txt
0 [main] ccdoc 603763 open_stackdumpfile: Dumping stack trace to CCDOC.EXE.stackdump
/cygdrive/c/CHIFUNG/BIN/mkscribdocs.sh: line 64: 603763 Segmentation fault (core dumped) ccdoc -db doc/webdocs/ccdoc.db -index
0 [main] ccdoc 94223007 open_stackdumpfile: Dumping stack trace to CCDOC.EXE.stackdump
/cygdrive/c/CHIFUNG/BIN/mkscribdocs.sh: line 70: 94223007 Segmentation fault (core dumped) ccdoc -db doc/webdocs/ccdoc.db
-html doc/webdocs/ -header root_hdr.txt -srcurl ../../ -root Scribble
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fix in r8.
Dangling prefix comment statements at the end of db caused an iteration to extend past the end of a collection in database.cc around line 400. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0021 ENTRY_TITLE: Report the version/revision information in the default trailer. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/08/28 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/28 ENTRY_REPORTED_BEGIN: Report the version/revision information in the default trailer. This makes error reporting easier. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r8. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0022 ENTRY_TITLE: Centralize the version/revision information. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/08/28 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/28 ENTRY_REPORTED_BEGIN: Currently there are several references to "0.8" in phase3_html.c and in switches.cc. These should be centralized to make maintenance easier. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r8.
All references to version are now funneled through ccdoc::switches::version(). ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0023 ENTRY_TITLE: DOC++ vs ccdoc ENTRY_REPORTED_BY: Bragas ENTRY_REPORTED_ON: 2001/08/28 ENTRY_STATUS: CLOSED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/18 ENTRY_REPORTED_BEGIN: I'm wondering if you couldn't merge your project with DOC++. They look very similar and joining your efforts could help creating even better tool (best in its category ?).
I'm not DOC++ developer and this is just my private impression.
As far as I am concerned, the DOC++ developers (or anyone else for that matter) are free to take the ccdoc code and incorporate it into their system(s). I would be happy to support them in that effort. Furthermore, if they showed sufficient commitment to my goal of improving documentation in the C++ community for free, I would be happy to hand it over completely. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Surveyed current ccdoc users and asked whether they wanted this project merged with Doxygen or DOC++. All (100%) of the respondents said no because they felt that the tools served different needs. Ccdoc was viewed as an interface documentation tool and doxygen was viewed as a source code documentation tool. In many cases, folks used both. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0024 ENTRY_TITLE: Missing comma separator in @see output for the alternative URL form. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/08/28 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/28 ENTRY_REPORTED_BEGIN: The @see directive does not correct separate entries with commas if there an alternative form URL specified. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r9.
Forgot to output leading comma's for the alternative URL form of the @see directive. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0025 ENTRY_TITLE: Undocumented 0.7a feature for explicitly specifying @pkgdoc URLs is no longer supported. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/08/28 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/28 ENTRY_REPORTED_BEGIN: In version 0.7a, ccdoc supported an undocumented syntax for the @pkgdoc directive that allowed the user to specify a custom URL. The syntax looked like this:
/** @pkgdoc A.B ../A_B.html*/
Ccdoc would create the link using the standard package name and substitute in
the specified link. We are using that feature extensively. Will it be supported
in v0.8?
Added support for the two argument form of @pkgdoc and documented it in the on-line help and in the users guide. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0026 ENTRY_TITLE: Provide an expanded package contents view. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/08/29 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/29 ENTRY_REPORTED_BEGIN: This would replace the tree view in the previous version. Basically it would expand the contents path hierarchically (the same way it is done in the class summary) with the type, scope and short description.
I implemented the -[no]rpthpc. It reports the hierarchical class structure of each package. I liked the format so much that I made it the default.
I also added the -[no]rptcsd to allow you to control whether details are reported on the class summary page (report class summary details). ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0027 ENTRY_TITLE: Provide a way to turn off package source. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/08/29 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/29 ENTRY_REPORTED_BEGIN: The source entry on the package page always points to a ccdoc db file. This is not useful to casual readers.
Completely removed source statements for packages. The db reference does not make sense for users. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0028 ENTRY_TITLE: Provide a way to turn off the default author on the package page. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/08/29 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/29 ENTRY_REPORTED_BEGIN: Reporting unascribed as the default package author when the author is not specified is not really useful in our environment. Can I turn this default handling off?
Added the -[no]rptdpa switches. The default is -norptdpa. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0029 ENTRY_TITLE: Provide a way to turn off the default version on the package page. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/08/29 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/29 ENTRY_REPORTED_BEGIN: Reporting unknown as the default package version when the version is not specified is not really useful in our environment. Can I turn this default handling off?
Added the -[no]rptdpa switches. The default is -norptdpv. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0030 ENTRY_TITLE: Provide a way to turn off default description comments on the page. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/08/29 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/29 ENTRY_REPORTED_BEGIN: Reporting undocumented as the default package comment when the comment is not specified is not really useful in our environment. Can I turn this default handling off?
Added the -[no]rptdpd switches, made -norptdpd the default. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0031 ENTRY_TITLE: Provide a way to change the "package" string in the title on package pages. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/08/29 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/29 ENTRY_REPORTED_BEGIN: In our environment packages have different semantics, some of them are subsystems, others are packages and still others are layers. I would like to be able to change the title on the package page to reflect this.
Added the @pkgdoctid directive. See the on-line help or the users guide for additional details. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0032 ENTRY_TITLE: Users guide and on-help report than -norptpub is the default. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/08/29 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/29 ENTRY_REPORTED_BEGIN: Minor documentation nit, help and the users guide report that -norptpub is the default when the real default is -rptpub. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r10. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0033 ENTRY_TITLE: This construct is incorrectly recognized as a struct: void foo(struct bar* spam). ENTRY_REPORTED_BY: Wawszczak, Jarosław ENTRY_REPORTED_ON: 2001/08/30 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/08/30 ENTRY_REPORTED_BEGIN: This construct is incorrectly recognized as a struct: void foo(struct bar* spam). ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r10.
This occurred because the parsing heuristic didn't recognize the the struct declaration was inside an argument list. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0034 ENTRY_TITLE: For -rpthpc only show packages. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/09/02 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/04 ENTRY_REPORTED_BEGIN: I have one more request.... I'd like the 'hierarchical contents' to show 'packages only', and not classes, namespaces, etc. Basically, I want to limit the depth to packages. The 'classes' link can still be the full glory. Without this, the current contents with classes is too detailed for the top-level index. With my mod, it will be a direct replacement of the old 'tree'.
Added a quick reference guide at the top level. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0036 ENTRY_TITLE: Comments get lost. ENTRY_REPORTED_BY: Yuriy Stul ENTRY_REPORTED_ON: 2001/09/03 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/04 ENTRY_REPORTED_BEGIN: Comments get lost under the following conditions:
/**
* This is a dangling comment at the head of a file.
*/
/**
* This is a class comment that gets lost.
*/
class A {};
You can work around this problem by changing the initial comment to
a non-ccdoc style comment as follows:
/*
* This is a dangling comment at the head of a file.
*/
/**
* This is a class comment that gets lost.
*/
class A {};
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r12.
This problem occurred because the database reader assumed that comments could not exist in sequential records. The fix was in database.cc around line 399. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0037 ENTRY_TITLE: Change the asterisk column header to "Inherited From". ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/09/05 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/05 ENTRY_REPORTED_BEGIN: Change the asterisk column header to "Inherited From". The asterisk is not very meaningful. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r13. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0038 ENTRY_TITLE: Change the Path column header to "Entity". ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/09/05 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/05 ENTRY_REPORTED_BEGIN: Change the asterisk column header to "Inherited From". The asterisk is not very meaningful. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r13. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0039 ENTRY_TITLE: Performance problem in phase3. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/09/05 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/07 ENTRY_REPORTED_BEGIN: The new version of ccdoc seems to generate HTML files much slower than v0.7a. In our tests it was running about 10X slower. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r14.
Found an O(N^2) complexity algorithm in databases::get_stmt_no_pkgs(). Converted the algorithm to O( n log(n) ). The test case performance improved from 49 minutes to 30 seconds but the real case is still slow.
Found another O(N^2) complexity algorithm in databases::get_stmt_no_pkgs() related to tree traversal for id lookups. Converted the algorithm to O( n log(n) ). The real test case performance improved from 18 hours to 12 minutes.
Made the generated HTML slightly more compact. The resulting output in the browers is identical.
Summary of performance improvements:
Summary of other miscellanous changes:
The make_class_contents() method in phase3_html.cc was not verifying that the inherited methods were accessible. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0041 ENTRY_TITLE: Improve 0039 performance enhancement. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/09/07 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/08 ENTRY_REPORTED_BEGIN: The following changes would improve the issue 0039 performance fix even more:
It seems to be reporting a link and link text in the contents of a package as the value (it is a static global variable). There are two and the comments are generated correctly, but the link goes to the first one.
Here is the snippet of code with comments:
And this is the resulting html snippit:typedef int Bool; #define true 1 #define false 0 /** * Global boolean value to signal if an account is loaded. * * @author Paul Kohler */ static Bool g_bAccountInit = false; /** * Global boolean value to signal if the connection has been established. * * @author Paul Kohler * @see IConnect */ static Bool g_bConnected = false;
false(href="ccdoc.lib.libivr.false.var.html") variable public Global boolean value to signal if the ICSS connection has been established. false(href="ccdoc.lib.libivr.false.var.html") variable public Global boolean value to signal if an account is loaded.
The variable parsing heuristic was not correctly recognizing '=' as an id separator in some cases. Look for "Issue 0042" in a comment in phase1_parser.cc. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0043 ENTRY_TITLE: Old hash mark syntax for @see does not work. ENTRY_REPORTED_BY: Lou Sanchez-Chopitea ENTRY_REPORTED_ON: 2001/09/10 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/10 ENTRY_REPORTED_BEGIN: The old @see #foo syntax no longer works.
This was an oversight. I inadvertently deprecated it. It will be fixed in the next release. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r15. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0044 ENTRY_TITLE: The summary report could use a little more indenting. ENTRY_REPORTED_BY: Lou Sanchez-Chopitea ENTRY_REPORTED_ON: 2001/09/10 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/10 ENTRY_REPORTED_BEGIN: It would be nice to be able to control the indenting level for reporting hierarchical objects on the summary page..
Perhaps I could add a switch that would allow you to control the level of indenting. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r15.
Added the -rptcsi (Class Summary Indent) switch. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0045 ENTRY_TITLE: The variable width font in the Code: section is not as readable as the fixed width font used in 0.7a. ENTRY_REPORTED_BY: Lou Sanchez-Chopitea ENTRY_REPORTED_ON: 2001/09/10 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/11 ENTRY_REPORTED_BEGIN: The variable width font in the Code: section is not as readable as the fixed width font used in 0.7a.
I have no opinion one way or the other.
I will add a switch that allows the user to control this. of indenting. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r17.
Added the -rptfwcf (Fixed Width Code Font) switch.
I originally reported this as fixed in r15 but the generate HTML still seemed to be proportional so I changed the generated output. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0046 ENTRY_TITLE: The @see index doesn't seem to work. ENTRY_REPORTED_BY: Chris Meuzelaar ENTRY_REPORTED_ON: 2001/09/10 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/10 ENTRY_REPORTED_BEGIN: ccdoc fails to generate links for @see directives when generating documentation for a large number of files / classes when multiple instances of the function exist in a class (see below). The sample below will generate HTML with the link back to the first function CSampleTray::AddSampleAt 0, but not the other two. Attached are the batch files used to generate the documentation as well as the database file. I am generating documentation using ~500 files.
Here is a bit of sample code that doesn't work.
class A {
public:
/**
* Do stuff.
* @see A::B 1
* @see A::B 2
*/
void B();
void B(int);
void B(double);
};
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r16.
I thought that this bug was important enough to warrant a re-release. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0047 ENTRY_TITLE: The indenting on the contents page should be the same as the summary report. ENTRY_REPORTED_BY: Lou Sanchez-Chopitea ENTRY_REPORTED_ON: 2001/09/11 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/11 ENTRY_REPORTED_BEGIN: The indenting on the contents page should be the same as the summary report. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r17.
Modified the -rptcsi (Class Summary Indent) switch to also change the contents indenting. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0048 ENTRY_TITLE: Derived classes in namespaces are not recognized correctly. ENTRY_REPORTED_BY: Dmitry A.Steklenev ENTRY_REPORTED_ON: 2001/09/11 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/11 ENTRY_REPORTED_BEGIN: In the following example, the HTML for class_two does not correctly link to class_one.
namespace ccdoc_test {
class class_one {
};
class class_two : public class_one {
};
}
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r17.
Partially scoped names were not being expanded in the extends clause processing in phase3_html.cc. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0049 ENTRY_TITLE: Non-template class method instantiation not recognized correctly. ENTRY_REPORTED_BY: Dmitry A.Steklenev ENTRY_REPORTED_ON: 2001/09/11 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/17 ENTRY_REPORTED_BEGIN: In the following example, some_method() is reported as a separate function.
namespace ccdoc_test {
class class_one {
public:
int some_method();
};
inline int class_one::some_method()
{
return 0;
}
}
namespace ccdoc_test {
class class_one {
public:
int some_method();
};
#if !defined( __ccdoc__ )
inline int class_one::some_method()
{
return 0;
}
#endif
}
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r33 and verified by test 69.
This was fixed as part of the issue 0149 work. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0050 ENTRY_TITLE: Ccdoc fails to parse methods with fct args properly. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/09/11 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/11 ENTRY_REPORTED_BEGIN: The following example is reported incorrectly as a variable of name fp:
class A {
public:
int fct2( void ( * fp ) ( ) ) ;
};
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r17.
This bug resulted from the phase_parser not being able to correctly identify methods under certain conditions. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0051 ENTRY_TITLE: External class method implementations with default arguments are not correctly recognized. ENTRY_REPORTED_BY: Lou Sanchez-Chopitea ENTRY_REPORTED_ON: 2001/09/17 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/17 ENTRY_REPORTED_BEGIN: External class method implementations with default arguments are not correctly recognized. Here is an example:
class A {
public:
void fct(int x=0);
};
void A::fct(int x)
{
}
Ccdoc treats the implementation as a separate function.
Because there are workarounds, this a low priority bug. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r33 and verified by test 69.
Workaround #1 Workaround #2 class A { public: void fct(int x=0); }; #if !defined(__ccdoc__) void A::fct(int x) { } #endifclass A { public: void fct(int x /**@#-*/ =0 /**@#+*/ ); }; void A::fct(int x) { }
This was fixed as part of the issue 0149 work. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0052 ENTRY_TITLE: Backslash character definition ('\\') confuses the ccdoc scanner. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/09/18 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/18 ENTRY_REPORTED_BEGIN: The following definition confuses the ccdoc parser.
const char X = '\\';
There was a scanner problem for detecting the ends of character constants in some cases. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0053 ENTRY_TITLE: No line numbers for internal comment parsing warnings. ENTRY_REPORTED_BY: Ray Rizzuto ENTRY_REPORTED_ON: 2001/09/18 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/21 ENTRY_REPORTED_BEGIN: Error during parse of header file with a package comment:
ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r19. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0054 ENTRY_TITLE: Serious problem with typedef of struct ENTRY_REPORTED_BY: Dékány Dániel ENTRY_REPORTED_ON: 2001/09/20 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/21 ENTRY_REPORTED_BEGIN: The following struct typedef conduses ccdoc.WARNING: Internal comment parsing error. Unexpected EOF for ''. This comment will be ignored. WARNING: Internal comment parsing error. Unexpected EOF for ''. This comment will be ignored.
typedef struct _TEST {
int a;
int b;
} TEST1;
It generates the following documentation:
package ccdocTest
Contents
Entity Type Scope Short Description
a typedef public undocumented
b variable public undocumented
You might be able to workaround this problem as follows:
struct _TEST {
int a;
int b;
};
typedef _TEST TEST1;
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r19.
The ccdoc::phase1::parser::parse_typedef() method did not correctly recognize embedded structures. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0055 ENTRY_TITLE: ccdoc generates default special members for C style structs ENTRY_REPORTED_BY: Dékány Dániel ENTRY_REPORTED_ON: 2001/09/20 ENTRY_STATUS: CLOSED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/20 ENTRY_REPORTED_BEGIN: Ccdoc should never create default special members (constructor, destructor, operator=) for structs in C programs, since there is no such thing in C.
Perhaps this problem can be solved in another way by allowing you to specify -nocdsm in phase1 for certain files. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Client has a workaround. I will re-open this issue if I get additional requests. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0056 ENTRY_TITLE: Incorrect line numbers reported in warning messages. ENTRY_REPORTED_BY: Dékány Dániel ENTRY_REPORTED_ON: 2001/09/20 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/18 ENTRY_REPORTED_BEGIN: Try this:
---[ccdoctest.h]--- /** @pkg ccdocTest */ int x;Ccdoc will say, that x is in ccdoctest.h:4, but it is in ccdoctest.h:3. It seems to me, that the extra 1 line number is related somehow to the ccdoc comment. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r34, verified by test 13.
This was extremely tricky because the line numbers were correct most of the time so it was hard to find a test case. The only time they were incorrect were when the first line of the file had a ccdoc directive that required a push/pop of the tokens that included the new line. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0057 ENTRY_TITLE: "Undocumented" as default short desc. in tables is confusing. ENTRY_REPORTED_BY: Dékány Dániel ENTRY_REPORTED_ON: 2001/09/20 ENTRY_STATUS: CLOSED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/21 ENTRY_REPORTED_BEGIN: "Undocumented", the default short description can be confusing in HTML tables when most entries have a short description, since "Undocumented" entries can be easily overlooked. I think the solution is to give switch to override this text, then I can replace it with "-". ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Ccdoc already supports this. The resolution is described in detail in issue 0058. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0058 ENTRY_TITLE: Provide a way to turn off the default author and version for packages. ENTRY_REPORTED_BY: Dékány Dániel ENTRY_REPORTED_ON: 2001/09/20 ENTRY_STATUS: CLOSED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/21 ENTRY_REPORTED_BEGIN: Reporting unascribed as the default package author and reporting version as unknown is not so useful in some projects, since we do it on package level. Perhaps -[no]rptdnpa and -[no]rptdnpd, where np is for non-package entity? ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Ccdoc provides a mechanism for turning off the default author and version for any package by allowing the client to specify them explicitly. The technique for doing this is to create a text file with entries for each property. That file could look something like this:
This file would be added to the phase 1 flow as follows:// Create empty entries for packages to avoid using the defaults. /** * * @author * @version * @pkgdoc PKG1 */ /** * * @author * @version * @pkgdoc PKG1::SUBPKG1 */ /** * * @author * @version * @pkgdoc PKG1::SUBPKG2 */
ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0059 ENTRY_TITLE: Validate ccdoc generated HTML. ENTRY_REPORTED_BY: Dékány Dániel ENTRY_REPORTED_ON: 2001/09/20 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/20 ENTRY_REPORTED_BEGIN: You can use the verifier at http://validator.w3.org to find errors in ccdoc generated output.ccdoc -db $CCDOC_DATABASE pkgdoc.txt
This would be much better in many cases:CONTENTS Entry Type Scope Shor Description b function ... c macro ... d function ... e package ... f macro ...
CONTENTS Packages Entry Type Scope Short Description e package ... Functions Entry Type Scope Short Description b function ... d function ... Macros Entry Type Scope Short Description c macro ... f macro ...
A switch like -[no]grpbytype? I don't hink so, cos it's not enough generic (E.g. maybe an alphabetical index is required too). I think the real solution is to allow user to specify any number of tables, with custom grouping and title. Something like this:
It would generate two tables: one with title "Contents" and with grouping by entity type, and below another table with title "Index" without grouping (just an alphabetiacal index, like the table in the current version). I put the grouping condition into parenthesis to allow multi-level grouping for future developmens, like -contents(type,functionality).$ ccdoc ... -table(type) "Contents" -table() "Index" ... (Note that I have skipped unimportant parts with ellipsis)
In this case there is only one statement but it is clear that it could be broken into the struct definition and the typedef. It might be possible to modify ccdoc with a heuristic that recognizes this construct and breaks it into two.
This is a low priority enhancement because this syntax is C artifact that is not necessary in C++ because the struct [name] declares [name] as a type. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0062 ENTRY_TITLE: Long entry name deforms HTML table. ENTRY_REPORTED_BY: Dékány Dániel ENTRY_REPORTED_ON: 2001/09/23 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: Just an idea: Long entry name stretches the first column too much. In the worst case it makes the "Contents" table unusable, since it stretches the first column so much that there will be no sufficient space for the short description.
A possible solution is to give a switch to specify the maximum entry name length, then break long names into two (or three, etc...) with an "\\<br>\\" in the HTML (to be more precise, with '\\</a><br>' + indent*' ' + '<a href="' + relpath + '">\\' (two backslash in the HTML output, not C escaping of one backslash...)) ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0063 ENTRY_TITLE: Allow user defined defaults for author, short description and version. ENTRY_REPORTED_BY: Dékány Dániel ENTRY_REPORTED_ON: 2001/09/23 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/24 ENTRY_REPORTED_BEGIN: Reporting unascribed as the default package author and reporting version as unknown is not so useful in some projects, since we do it on package level. Perhaps -[no]rptdnpa and -[no]rptdnpd, where np is for non-package entity?
I believe that the correct way to address the problem is to allow the user to specify their own default strings for author, version and short description types using new switches: -rptdefa, -rptdefsd and rptdefv. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r20.
Added:
Code Sample:
namespace nms {
class some_class
{
public:
void some_class();
};
class some_other_class
{
/** Red, but not linked */
friend class some_class;
public:
void some_other_class();
};
}
In this particular case, I can modify ccdoc to try looking in the parents namespace it the id is not found. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r21.
Modified ccdoc:phase3::html::write_friends_info to look in the namespace of the parent to resolve unresolved names. Unfortunately this will not work for all cases. The best solution is to always fully qualify scoped names. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0065 ENTRY_TITLE: Links from documentation in namespaces are not recognized correctly. ENTRY_REPORTED_BY: Dmitry A.Steklenev ENTRY_REPORTED_ON: 2001/09/25 ENTRY_STATUS: CLOSED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/25 ENTRY_REPORTED_BEGIN: Code Sample:
namespace nms {
class some_class
{
public:
void some_class();
};
/**
* Some description
*
* Red, but not linked:
* @$ some_class
* and "see also" have same problem.
* @see some_class
*/
class some_other_class
{
public:
void some_other_class();
};
}
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Ccdoc does not resolve scoped identifiers (because it does not process
included header files that may contain "using" statements). This means that
the user must be explicitly specify the full scope of an identifier
as shown below.
namespace nms {
class some_class
{
public:
void some_class();
};
/**
* Some description
*
* Red, but not linked:
* @$ nms::some_class
* and "see also" have same problem.
* @see nms::some_class
*/
class some_other_class
{
public:
void some_other_class();
};
}
I added fix that will resolve partially scoped references in most cases. See issue 0118 for details. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0066 ENTRY_TITLE: The <sstream> header is not needed in phase3_html.cc. ENTRY_REPORTED_BY: Dmitry A.Steklenev ENTRY_REPORTED_ON: 2001/09/25 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/09/27 ENTRY_REPORTED_BEGIN: The <sstream> header is not needed in phase3_html.cc. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r21.r
You are absolutely correct. The <sstream> header is not needed.
Earlier I was using stringstreams but had a porting problem and had to remove them. When I did that I neglected to remove the header file.
Thanks for pointing it out. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0067 ENTRY_TITLE: Ccdoc does not correctly reset the access specifiers for methods in base classes that are derived as protected or private. ENTRY_REPORTED_BY: Dmitry A.Steklenev ENTRY_REPORTED_ON: 2001/10/01 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: Code Sample:
class some_base_class
{
public:
/** Must be reported as "protected" in some_class_01 and
* as "private" in some_class_02, but reported as "public" */
void some_public_method();
protected:
/** Must be reported as "protected" in some_class_01 and
* as "private" in some_class_02, but reported as "protected" */
void some_protected_method();
};
class some_class_01 : protected some_base_class {};
class some_class_02 : private some_base_class {};
This may be difficult for many users. I will look into changing this behavior or making it optional via a switch. Meanwhile, if you don't want the users to see the methods, you can work around this as follows:
class some_base_class
{
public:
/** Must be reported as "protected" in some_class_01 and
* as "private" in some_class_02, but reported as "public" */
void some_public_method();
protected:
/** Must be reported as "protected" in some_class_01 and
* as "private" in some_class_02, but reported as "protected" */
void some_protected_method();
};
class some_class_01
#if !defined(__ccdoc__)
: protected some_base_class
#endif
{};
class some_class_02
#if !defined(__ccdoc__)
: private some_base_class
#endif
{};
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0068
ENTRY_TITLE: Core dump on solaris in the index phase.
ENTRY_REPORTED_BY: John Rostykus
ENTRY_REPORTED_ON: 2001/10/01
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2001/10/01
ENTRY_REPORTED_BEGIN:
Under certain conditions a core dump is generated during the index
phase for ccdoc r20.
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r21.
This problem occurred because phase2 was deleting statements without removing them from the load map in the database object. This caused a bad pointer to be referenced under certain conditions. This bad pointer caused the core dump on solaris. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0069 ENTRY_TITLE: Cannot handle construct: "Float operator/(const Float & rval)" ENTRY_REPORTED_BY: Gustavsson Mĺrten ENTRY_REPORTED_ON: 2001/10/04 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/08 ENTRY_REPORTED_BEGIN: Ccdoc cannot handle the construct:
You can work around this problem by inserting a space after the forward slash as follows:Float operator/(const Float & rval)
ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r22.Float operator/ (const Float & rval)
There was a bug in ccdoc::phase1::scanner::scan_token() for processing forward slash tokens. When the scanner looked ahead and failed, it did not reset the next character properly. The problem was easily fixed by resetting the next character properly. This occurred around line 506 in phase1_scanner.cc 1.8. Test 29 was added to verify the fix. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0070 ENTRY_TITLE: Users guide documentation error. ENTRY_REPORTED_BY: Gustavsson Mĺrten ENTRY_REPORTED_ON: 2001/10/04 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/08 ENTRY_REPORTED_BEGIN: By the way, the following section in the user documentation contains an error:
It should be:Directive: /**@#-*/, /**@#+*/
Turns ccdoc token processing off and on. They are useful for turn off processing for tokens on a line. For whole lines, use the __ccdoc__ compiler pragma as shown in the example below:// Turn off sections of code using __ccdoc__ #if defined(__ccdoc__) // Ignore stuff until the end #endif
ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r22. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0071 ENTRY_TITLE: Operator keyword duplicated for method operators. ENTRY_REPORTED_BY: Lou Sanchez-Chopitea ENTRY_REPORTED_ON: 2001/10/08 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/08 ENTRY_REPORTED_BEGIN: Operator keyword duplicated for method operators. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: .#if !defined(__ccdoc__)
There was a bug in ccdoc::phase3::html::write_section_header for recognizing method operators for creating titles. It was not checking for statement::base::STMT_METHOD_OPERATOR when creating the title.
This is verified by test19. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0072 ENTRY_TITLE: New switch to turn off sorting of class entities. ENTRY_REPORTED_BY: Michael Wang ENTRY_REPORTED_ON: 2001/10/11 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/12 ENTRY_REPORTED_BEGIN: I would like to see a switch that allows the user to turn off output ordering/sorting so that the listing of methods/functions can be dictated by the programmer.
In keeping with the current naming convention this switch should be called something like -rptdsfce (disable sorting for class entities). ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r22.
Added two new switches: -rptsci and -norptsci to allow the user to control the sorting of class information. The default is -rptsci. If -norptcsi is specified the class contents and details are not sorted.
The idea is that programmers may wish to logically organize class information in a way that is more understandable than a simple sort. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0073 ENTRY_TITLE: Can't tell ccdoc to ignore a privately declared default constructor. ENTRY_REPORTED_BY: Lou Sanchez-Chopitea ENTRY_REPORTED_ON: 2001/10/11 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/12 ENTRY_REPORTED_BEGIN: When I create a class with a private default constructor ccdoc still reports it. Here is an example:
class A {
private:
A(); // ccdoc always reports this
public:
A(int);
};
I would like to see a directive that allows the user to turn off output generation for a specific entity. This would allow me to tell ccdoc to ignore the reporting of private default constructors on a per class basis.
Ccdoc must be fixed to not report private entities (even those generated by default) unless the -rptpri switch is specified. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r22.
Ccdoc will not report the methods and operators generated by default if they are privately declared. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0074 ENTRY_TITLE: Ccdoc cannot handle international character sets. ENTRY_REPORTED_BY: Dmitry A.Steklenev ENTRY_REPORTED_ON: 2001/10/12 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/12 ENTRY_REPORTED_BEGIN: CCDoc inserts a generated html meta variable:
Please, provide any command line arguments for switching to user specified charset.<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
I will add a switch that allows you to define your own character set in the next release (probably on Monday). ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r22.
Added a new switch called -rptctcs that allows users to specify the character set that they wish to use. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0075 ENTRY_TITLE: Ccdoc cannot handle function typedef of the form: typedef void fct(int arg1); ENTRY_REPORTED_BY: Dékány Dániel ENTRY_REPORTED_ON: 2001/10/13 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/14 ENTRY_REPORTED_BEGIN: Ccdoc cannot handle function typedef of the form: typedef void fct(int arg1);. It reports arg1 as the typedef rather than fct. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r23. Verified by test33.
The phase1_parser did not understand this syntax correctly. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0076 ENTRY_TITLE: Conflict between -meta switch and ccdoc defined meta variables. ENTRY_REPORTED_BY: Dékány Dániel ENTRY_REPORTED_ON: 2001/10/14 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/16 ENTRY_REPORTED_BEGIN: There is a conflict between the -meta switch and the fact that ccdoc automatically creates some meta-variables for HTML 4.01 compliance because it does not allow users to override or change the ccdoc generated meta variables.
If user gives a -meta switch, ccdoc must *not* generate standard meta elements, like http-equiv-s or keyword-s. Generating ccdoc_... metas is OK.
Also added a warning if both -meta and -rptctcs are specified. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0077 ENTRY_TITLE: Ccdoc incorrectly identifies a class attribute name. ENTRY_REPORTED_BY: Clint Bauer ENTRY_REPORTED_ON: 2001/10/18 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/19 ENTRY_REPORTED_BEGIN: Ccdoc identifies the attribute name as BAR instead of m_foo:
class A {
public:
char m_foo[BAR];
};
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r23. Verified by test35.
The variable id recognizer in ccdoc::phase1::parser::parse_var_or_fct (phase1_parser.c) was confused by the square brackets. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0078 ENTRY_TITLE: Ccdoc incorrectly identifies a variable as a function. ENTRY_REPORTED_BY: Clint Bauer ENTRY_REPORTED_ON: 2001/10/18 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/19 ENTRY_REPORTED_BEGIN: Ccdoc incorrectly identifies a variable as a function in the following case:
If the parentheses are removed, ccdoc correctly identifies it as a variable. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r23. Verified by test36.const long is_not_a_function = (1+1);
The function type recognizer in ccdoc::phase1::parser::parse_var_or_fct (phase1_parser.c) was confused by the presence of parentheses after an equals sign. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0079 ENTRY_TITLE: platform.pl cannot handle embedded parentheses. ENTRY_REPORTED_BY: Ben Zoe ENTRY_REPORTED_ON: 2001/10/23 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Ben Zoe ENTRY_RESOLVED_ON: 2001/10/24 ENTRY_REPORTED_BEGIN: I use cygwin and it attaches a trailer build number in parentheses which is an illegal path name. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r24.
Ben added a line to filter out parentheses to platform.pl. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0080 ENTRY_TITLE: Trim trailing spaces in path name for get_stmt_no_pkgs(). ENTRY_REPORTED_BY: Ben Zoe ENTRY_REPORTED_ON: 2001/10/23 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Ben Zoe ENTRY_RESOLVED_ON: 2001/10/24 ENTRY_REPORTED_BEGIN: If the path name has trailing spaces, the get_stmt_no_pkgs() may fail unexpectedly. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r24. Verified by test #38.
Ben added code that will trim trailing white space. I modified it slightly to include the \r white space character for PCs.
It looks like this problem can be tickled by @links with trailing spaces. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0081 ENTRY_TITLE: Multi-space blank lines between SHORT and LONG descriptions cause them to merge. ENTRY_REPORTED_BY: Ben Zoe ENTRY_REPORTED_ON: 2001/10/23 ENTRY_STATUS: CLOSED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/24 ENTRY_REPORTED_BEGIN: Multi-space blank lines between SHORT and LONG descriptions cause them to merge. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: The current ccdoc behavior is correct. Blanks are used to indicates lines in HTML <pre> clauses. The fix that was described (testing for lines that contain w/s) would fail to properly parse the following case:
/** * This is a short description with some code in it: ** // Whitespace test * ** * This is the start of the long description. */
I am not sure that this resolution is correct because the ccdoc short description does not conform to the javadoc specification. See issue 0089 for details. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0082 ENTRY_TITLE: Add the @since directive per the javadoc 1.2 specification. ENTRY_REPORTED_BY: Ben Zoe ENTRY_REPORTED_ON: 2001/10/23 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Ben Zoe & Joe Linoff ENTRY_RESOLVED_ON: 2001/10/25 ENTRY_REPORTED_BEGIN: Add the @since directive per the javadoc specification. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r24. Verified by test #39.
This change invalidates old ccdoc db files. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0083 ENTRY_TITLE: Modify ccdoc to create a separate links page (as was done in v0.7a). ENTRY_REPORTED_BY: Ben Zoe ENTRY_REPORTED_ON: 2001/10/23 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/26 ENTRY_REPORTED_BEGIN: When ccdoc see's multiple links to the same entities it creates a list. In ccdoc v0.7a it created a separate page with the links. I like that approach better.
Modified ccdoc to handle multiple links on a separate page. It is much easier to read than the previous version. The main work was done in the ccdoc::phase3::html::new write_links() method. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0084 ENTRY_TITLE: Alias @exception to @throws per javadoc 1.2. ENTRY_REPORTED_BY: Ben Zoe ENTRY_REPORTED_ON: 2001/10/23 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Ben Zoe & Joe Linoff ENTRY_RESOLVED_ON: 2001/10/25 ENTRY_REPORTED_BEGIN: Alias @exception to @throws per javadoc 1.2. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r24. Verified by test #39. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0085 ENTRY_TITLE: Ccdoc not recognize default constructors with member initialization. ENTRY_REPORTED_BY: Ben Zoe ENTRY_REPORTED_ON: 2001/10/23 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/25 ENTRY_REPORTED_BEGIN: Ccdoc not recognize default constructors with member initialization as shown in the following example:
class A {
public:
A() : x(0), y(0) {}
private:
int x,y;
};
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r24. Verified by test #42.
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0086
ENTRY_TITLE: Comment form to handle single line suffix comments.
ENTRY_REPORTED_BY: Ben Zoe
ENTRY_REPORTED_ON: 2001/10/24
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Ben Zoe
ENTRY_RESOLVED_ON: 2001/11/25
ENTRY_REPORTED_BEGIN:
One another topic, I was looking for a way to do a one line ccdoc
comment for this type of suffix documentation:
class A
{
public:
/** An int variable */
int intVar;
/** A long variable */
long longVar;
};
as
class A
{
public:
int intVar; //@- An int variable
long longVar; //@- A long Var
};
This is not a priority item. It probably has limited application. But I
personally have many files where a single short comment line to the
right of the documented item makes the file easier to read. I tried
finding a place to do this but stopped when it took too much time. I'll
keep at it when I some spare time.
Ben did all of the work to implement the changes to phase1_scanner.cc. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0087 ENTRY_TITLE: Exceptions were not reported as links. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/10/25 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/25 ENTRY_REPORTED_BEGIN: Ccdoc was not reporting links to exceptions in @exception and @throws directives. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r24. Verified by test #39.
The code for resolving the exception class was not implemented in ccdoc::phase3::html::write_ccdoc_exception_directive_info() (phase3_html.c). ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0088 ENTRY_TITLE: An extra invalid link reported for @link references to classes. ENTRY_REPORTED_BY: Ben Zoe ENTRY_REPORTED_ON: 2001/10/23 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/25 ENTRY_REPORTED_BEGIN: An extra invalid link reported for @link references to classes. The following code fragment demonstrates the problem:
class A {
};
class B {
public:
/**
* Comment.
* @link A
*/
void fct();
};
The link will be reported as "link: A {A}" where the second link is invalid.
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r24. Verified by test #39.
The STMT_CLASS_END statement (a dummy) was being reported as an invalid link. This was corrected by adding ccdoc::phase3::html::find_and_write_links() to detect and ignore end of class statements as well other send statements for structs, unions and namespaces. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0089 ENTRY_TITLE: Ccdoc short description syntax is not compliant with javadoc. ENTRY_REPORTED_BY: Mĺrten Gustavsson ENTRY_REPORTED_ON: 2001/10/25 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/25 ENTRY_REPORTED_BEGIN: From the javadoc manual http://java.sun.com/j2se/javadoc/writingdoccomments/index.html#generalform:
First SentenceCcdoc doesn´t handle first sentences according to javadoc standard. I believe the summary is created from the first paragraph.
The first sentence of each doc comment should be a summary sentence, containing a concise but complete description of the API item. This means the first sentence of each member, class, interface or package description. The Javadoc tool copies this first sentence to the appropriate member, class/interface or package summary. This makes it important to write crisp and informative initial sentences that can stand on their own.This sentence ends at the first period that is followed by a blank, tab, or line terminator, or at the first tag (as defined below). For example, this first sentence ends at "Prof.":
/** * This is a simulation of Prof. Knuth's MIX computer. */
I think that there is a point in following the javadoc specification as close as possible, the documentation is for free.
I feel that the best solution is to modify ccdoc to conform to the javadoc standard but provide a switch that allows the current behavior. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r24. Verified by test #40.
Changed the default behavior of ccdoc to conform to javadoc and added two phase 1 switches: -[no]jdsds. The -jdsds (JavaDoc Short Description Syntax) switch enables the javadoc behavior and is the default. The -nojdsds switch enables the old behavior for backward compatibility.
Beware! This change will affect your output. Use the -nojdsds switch to enable the old behavior. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0090 ENTRY_TITLE: Support the {@link ...} javadoc syntax. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/10/25 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/26 ENTRY_REPORTED_BEGIN: Ccdoc does not support inline link references. It should be enhanced to support the {@link ...} directive.
As an interesting aside, the @link directive in ccdoc was developed independently of javadoc. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r24. Verified in test #44. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0091 ENTRY_TITLE: Constant values are not hyperlinked in the documentation for function arguments. ENTRY_REPORTED_BY: David McBride ENTRY_REPORTED_ON: 2001/10/25 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/25 ENTRY_REPORTED_BEGIN: I have a suggestion. I've noticed that if an enum is a parameter in a funtion, then a link gets created back to the enum definition. That's perfect. However, it doesn't seem that the same is true for constants defined either in the global or the class namespace (unless I'm missing something). It would be handy if ccdoc treated constants the same as enums in this regard.
Example:
enum myEnum {..........};
const double myConstant = 30.0;
class myClass
{
public:
void myFunct(myEnum fred, double ethel = myConstant);
}
ccdoc creates a link in myFunct at "myEnum" back to the definition of
"myEnum". However, no link is created in myFunct at "myConstant".
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r24. Verified by test #41.
The ccdoc::phase3::html::write_code_subsection_token() method was not looking for variables or functions. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0092 ENTRY_TITLE: The @link syntax does not support the # scoping operator. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/10/26 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/10/26 ENTRY_REPORTED_BEGIN: Ccdoc does not support the '#' scoping operator in @link directives. It works correctly for @see directives. The example below illustrates the problem.
class A {
public:
void fct();
/**
* {@link #fct foo}
* @link #fct
* @see #fct
*/
void fct1();
};
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r24. Verified in test #44.
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0093
ENTRY_TITLE: Add javadoc style doc comment inheritance.
ENTRY_REPORTED_BY: Mĺrten Gustavsson
ENTRY_REPORTED_ON: 2001/11/01
ENTRY_STATUS: OPEN
ENTRY_RESOLVED_BY:
ENTRY_RESOLVED_ON:
ENTRY_REPORTED_BEGIN:
Javadoc has quite an advanced feature for doc comment inheritance, see
http://java.sun.com/j2se/1.4/docs/tooldocs/win32/javadoc.html#inheritingcomments.
I would very much appreciate doc comment inheritance, prefferably according to javadoc 1.4. Javadoc 1.3 (http://java.sun.com/j2se/1.3/docs/tooldocs/win32/javadoc.html#inheritingcomments) is not so advanced but would be quite useful also.
I believe that the javadoc rules says that a method must override a method in the ancestor class to reuse a doc comment. Doc comments cannot be reused for constructors, destructors etc.
I find it quite tedious to write doc comments for constructors and destructors. It would save a lot of typing if one could reuse doc comments beween them.
I can see that the implementation can be difficult, but there is another enhancement that could simplify a lot. Lets call it doc comment reuse of overloaded functions. There is no such thing in javadoc, but maybe there should be. I have read the javadoc spec over and over since I really believed there was such a rule, but there is no rule.
The result is that one only needs to document the first function in a group of overloaded functions.
If the original function has a doc comment the generated function will be placed after the original one, inheriting its doc comment using the rule for overloaded functions.
If the original function doesn't have a doc comment, the generated one will be placed before the original one, allowing the original one to inherit from the generated one using the rule for overloaded functions.
A code example:
class Base
{
/** Default construcor */
Base();
/**
* Copy constructor.
*
* @param src is the object to copy from
*/
Base(const Base &src);
/** Destructor */
~Base();
/**
* Assignment operator.
*
* @param src is the object to copy from
* @return a const reference to this object
*/
const Base& operator=(const Base& src);
} // class Base
class Derived : public Base
{
Derived(); // inherit doc comment
Derived(const Derived &src); // inherit doc comment
~Derived(); // inherit doc comment
const Derived& operator=(const Derived & src); // inherit doc comment
/**
* Strange copy constructor that slices the object.
*
* @param src is the base class object to copy from
*/
Derived(const Base &src);
} // class Derived
Theses classes will internally be represented as something like this:
class Base
{
/** Default construcor */
Base();
this(); // inherit doc comment from previous using overload rule
/**
* Copy constructor.
*
* @param src is the object to copy from
*/
Base(const Base &src);
this(const this &src);// inherit doc comment from previous using overload rule
/** Destructor */
~Base();
~this();// inherit doc comment from previous using overload rule
/**
* Assignment operator.
*
* @param src is the object to copy from
* @return a const reference to this object
*/
const Base& operator=(const Base& src);
const this& operator=(const this& src);// inherit doc comment from previous using overload rule
} // class Base
class Derived : public Base
{
this(const this &src); // this one will pick up doc comment from base class
Derived(); // inherit doc comment from previous using overload rule
this(const this &src); // this one will pick up doc comment from base class
Derived(const Derived &src); // inherit doc comment from previous using overload rule
~this(); // this one will pick up doc comment from base class
~Derived(); // inherit doc comment from previous using overload rule
const this& operator=(const this& src); // this one will pick up doc comment from base class
const Derived& operator=(const Derived & src); // inherit doc comment from previous using overload rule
/**
* Strange copy constructor that slices the object.
*
* @param src is the base class object to copy from
*/
Derived(const Base &src);
} // class Derived
The output looks like this:/** * Returns bug? * @returns Are extra * comma's inserted * at each * new line? */ void fct();
Are extra , comma's inserted , at each , new line?
class MyClass {
public:
/**
* Performs some useful function. blah blah blah.
* Some long, involved explanation that I don't
* want to repeat over and over again.
* @param someVar Some user defined variable.
* @returns Returns 0 if successful or not 0 otherwise.
*/
void myOverloadedFunct(A* someVar);
/**
@repeat
*/
void myOverloadedFunct(B* someVar);
/**
@repeat
*/
void myOverloadedFunct(C* someVar);
/**
@repeat
*/
void myOverloadedFunct(D* someVar);
/**
@repeat
*/
void myOverloadedFunct(E* someVar);
}
class A {
public:
void fct();
private:
void fct();
};
The private void fct shows up in the list.
ccdoc generates a default constructor entry if there is no default constructor defined. This is not according to C++ standard. I suggest that you change ccdoc to only generate documentation for implicitly defined default constructors if there are no constructors defined in the class.
Also if you call generated member functions "implicitly defined" this will improve understanding since it will adhere to C++ lingo. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r26. Verified by test 55.
Modified special member processing to recognize when any constructor is declared so that it can turn off the generation of the default constructor correctly. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0100 ENTRY_TITLE: If using switch -files with a wildcard expression the first file is not processed. ENTRY_REPORTED_BY: Mĺrten Gustavsson ENTRY_REPORTED_ON: 2001/11/09 ENTRY_STATUS: CLOSED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/11/12 ENTRY_REPORTED_BEGIN: I saw in the issue list that someone has a problem with gnuwin32. If using switch -files with a wildcard expression the first file is not processed. No problem if removing the -files switch. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: The first file is not really ignored because the ccdoc expects it to contain a list of files to process. When it doesn't find any files it goes on to the other entries on the command line. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0101 ENTRY_TITLE: Template problem. ENTRY_REPORTED_BY: Mĺrten Gustavsson ENTRY_REPORTED_ON: 2001/11/09 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/11/27 ENTRY_REPORTED_BEGIN: The following code
template<class VALUE_TYPE> class Prompt : public Component
{
public:
/**
* Reads from the instream to an object and returns a reference to it.
* First, a text defined in the constructor is written to the ostream defined
* in the constructor. Then the instream defined in the constructor is
* read to the internal object, using operator>>(). A reference to the
* internal object is returned.
*/
Input< ReturnPrimitiveInterface<const VALUE_TYPE&, const Event&> > in;
generates this output:
attribute const ? ^ < >
Undocumented
Source:
metaf/src/components/io/Prompt.h:52
Code:
public Input < ReturnPrimitiveInterface < const VALUE_TYPE const Event & > > in
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r26. Verified in test 54.
The variable recognition heuristic in phase1::parser::parse_var_or_fct did not correctly recognize template arguments. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0102 ENTRY_TITLE: Returned nested classes and types in namespaces are not recognized correctly. ENTRY_REPORTED_BY: Dmitry A.Steklenev ENTRY_REPORTED_ON: 2001/11/10 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/11/26 ENTRY_REPORTED_BEGIN: Returned nested classes and types in namespaces are not recognized correctly.
Code Sample:
namespace nms {
class some_class
{
public:
typedef int nested_type;
class nested_class {};
};
class another_class
{
public:
/** Linked correctly */
some_class get_some_class();
/** Not linked correctly */
some_class::nested_type get_nested_type();
/** Not linked correctly */
some_class::nested_class get_nested_type();
};
}
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r26. Verified by test 52.
See issue 0118 for more details. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0103 ENTRY_TITLE: Invalid processing of operator~(). ENTRY_REPORTED_BY: Dmitry A.Steklenev ENTRY_REPORTED_ON: 2001/11/10 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/11/12 ENTRY_REPORTED_BEGIN: Sample code:
class some_class
{
public:
some_class operator~();
};
CCDoc output:
ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r25. Verified by test #47.~ operator public undocumented
The ccdoc::phase1::parser::get_fct_id() method was being called for operators. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0104 ENTRY_TITLE: Invalid processing of operator&(). ENTRY_REPORTED_BY: Dmitry A.Steklenev ENTRY_REPORTED_ON: 2001/11/10 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/11/12 ENTRY_REPORTED_BEGIN: Sample code:
class some_class
{
public:
some_class operator&( const some_class& );
};
CCDoc output:
ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r25. Verified by test #47.some_classoperator & operator public undocumented
The ccdoc::phase1::parser::get_fct_id() method was being called for operators. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0105 ENTRY_TITLE: The javadoc @see tag defines the label part as optional. ENTRY_REPORTED_BY: Mĺrten Gustavsson ENTRY_REPORTED_ON: 2001/11/08 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: The javadoc @see tag defines the label part as optional: http://java.sun.com/j2se/1.4/docs/tooldocs/win32/javadoc.html#@see
If you decide to do the label optional, javadoc specifies how to generate the link: http://java.sun.com/j2se/1.4/docs/tooldocs/win32/javadoc.html#shortened
I don´t feel that this improvement is very important, but if you like the idea put it on the issue list and maybe implement it if there is time and no more important things to do. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0106 ENTRY_TITLE: The javadoc spec for @see allows unqualified names as well as qualified names. ENTRY_REPORTED_BY: Mĺrten Gustavsson ENTRY_REPORTED_ON: 2001/11/08 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: The javadoc spec for @see allows unqualified names as well as qualified names: http://java.sun.com/j2se/1.4/docs/tooldocs/win32/javadoc.html#seesearchorder
I believe that ccdoc requires a fully qualified c++ class name if the reference is to another class. The javadoc spec relies on rules including the java compiler and I don´t believe that following their search patterns will add that much.
However, it would be nice if the @see tag accepts valid c++ class names. Leaving out the namespace part if the referencing class belongs to the same namespace as the referenced class would be an improvement.
I don´t think this improvement is very important since namespaces are quite short compared to java packages. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0107 ENTRY_TITLE: javadoc # syntax is not supported in the {@link...} directive. ENTRY_REPORTED_BY: Mĺrten Gustavsson ENTRY_REPORTED_ON: 2001/11/08 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/18 ENTRY_REPORTED_BEGIN: It seems like it isn´t possible to link to a member function in another class. Example from my code: * {@link metaf::ComAbstractGetAdoRecord#get_one_record ComAbstractGetAdoRecord::get_one_record()}
Javadoc allows this construct: http://java.sun.com/j2se/1.4/docs/tooldocs/win32/javadoc.html#@see
The syntax would be quite useful.
It turns out that ccdoc does allow you to reference member functions but it doesn't use the javadoc syntax, you must use a '::' separator as shown in the following example.
class A {
public:
void fct();
};
class B {
public:
/**
* See the {@link A::fct fct} for additional details.
* @see A::fct
*/
void bar();
};
I completely agree that ccdoc should support the javadoc syntax and
will schedule this fix for the next release.
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r34.
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0108
ENTRY_TITLE: -rptdefasd, -rptdefsd and -rptdefv problems and suggestions
ENTRY_REPORTED_BY: Mĺrten Gustavsson
ENTRY_REPORTED_ON: 2001/11/08
ENTRY_STATUS: OPEN
ENTRY_RESOLVED_BY:
ENTRY_RESOLVED_ON:
ENTRY_REPORTED_BEGIN:
I am playing around with the switches -rptdefasd, -rptdefsd and
-rptdefv, the intention to mark the text red "Undocumented" in order to find missing documentation
and implicitly defined member funcions.
For the implicitly defined member funcions I want the text "Implicitly defined", and for things that I forgot to document "Undocumented", the switches used:
-rptdefasd "Implicitly defined" -rptdefsd "Undocumented" This causes the class documentation look just like I want. However, I have found the following errors:
And some suggestions:
If the programmers intention is for C++ to automatically generate a constructor, there should be something in the documentation telling this, prefferably as a doc comment. If so there should be no red warning text.
Since I have so many classes, I find it tedious to click through all genereated html to find problematic links. I suggest that you also print warning messages to stderr. Finding the problems will be much faster.
If you decide to implement warnings for dangerous C++ constructs I guess that your user community will come up with even more ideas if asked. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r34, verified by test 73.
Added warning messages for all conditions that trigger red in the HTML code. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0110 ENTRY_TITLE: Invalid processing of operator[]. ENTRY_REPORTED_BY: Lou Sanchez-Chopitea ENTRY_REPORTED_ON: 2001/11/16 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/11/12 ENTRY_REPORTED_BEGIN: Invalid processing of operator~[]. This is exactly the same problem as issues 0103 and 0104. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r25. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0111 ENTRY_TITLE: Multiple link file overflow problem. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/11/16 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/11/17 ENTRY_REPORTED_BEGIN: I ran across a problem where multiple links caused too many files to be generated. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r25.
The problem was fixed by re-using common link files. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0112 ENTRY_TITLE: Phase 3 exception for operator|. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/11/19 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/11/19 ENTRY_REPORTED_BEGIN: Operator | clauses cause an exception to be generated in phase3. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r25. Verified by test48. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0113 ENTRY_TITLE: Merge separate ccdoc databases. ENTRY_REPORTED_BY: John Rostykus ENTRY_REPORTED_ON: 2001/11/20 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: On a related note, I would like to talk to you about possibly supporting a feature in ccdoc to merge databases. We could really parallelize the process if we were not limited to having every ccdoc process write to the same database. For example, run the header processing on each of the 50 or so subsystems individually in parallel, run a (new) step to merge the resulting databases into 1, and then do the index/html steps. Does ccdoc have any capability like this? If not, what do you think of this idea? In general, long processes that are single threaded are what kill our build times, and ccdoc (the way we use it) at ~1 hour is in that camp.
Sample code:
class some_class
{
public:
some_class operator unsigned long();
}
CCDoc output:
ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r26. Verified by test49.operator unsignedlong operator public undocumented
There was a name munging heuristic in the phase1::parser that collapsed operators like [] and (). It wasn't checking for cast operator entries like "unsigned long". ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0115 ENTRY_TITLE: Mangled links for .\doc\ references on the command line. ENTRY_REPORTED_BY: Richard Cooper ENTRY_REPORTED_ON: 2001/11/22 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/11/27 ENTRY_REPORTED_BEGIN: Running ccdoc with the following command line produces documentation with mangled links. The documents are in .\doc but the hyperlinks link to .\doc\doc
The system works fine if I do the following but as this will be part of an automatic process I don't want to.ccdoc -db docs.db -html .\doc\ ..\Headers\*.h
It also works if I move the doc directory to the same level as the Headers directorycd .\doc ccdoc -db docs.db -html .\ ..\..\Headers\*.h
but not if I move it further upccdoc -db docs.db -html ..\doc\ ..\Headers\*.h
ccdoc -db docs.db -html ..\..\doc\ ..\Headers\*.h
You can fix the problems that you were having by replacing the backslashes with forward slashes. Ccdoc will figure out the correct DOS path and the URL will work.
I apologize for not making this clear in the users guide. I will update the information and close out this issue if this works for you. If I don't hear from you by tommorrow (Nov 27) I will assume that things are working for you. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Updated the documentation to reflect the requirement that -html path specifications must use forward slashes on all platforms (even DOS). ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0116 ENTRY_TITLE: The string constant "\\" isn't parsed correctly. ENTRY_REPORTED_BY: Chris Martin ENTRY_REPORTED_ON: 2001/11/25 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Chris Martin ENTRY_RESOLVED_ON: 2001/11/25 ENTRY_REPORTED_BEGIN: The string constant "\\" isn't parsed correctly; resulting in a spurious "unterminated string literal" warning.
Fix:
void ccdoc::phase1::scanner::get_string_literal(char* token,int max)
{
int ch = '"';
int pch = ch;
while( max>0 && (ch = get_char()) ) {
max--;
*token++ = ch;
if( ch == '"' && pch != '\\' ) {
*token = 0;
return;
}
#ifdef CHRIS
if (ch == '\\' && pch == '\\') // escaped escape
ch = 0;
#endif
pch = ch;
}
.
.
Incorporated the fix by Chris and added a verification test. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0117 ENTRY_TITLE: Class method reported as global function. ENTRY_REPORTED_BY: Chris Martin ENTRY_REPORTED_ON: 2001/11/25 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Chris Martin ENTRY_RESOLVED_ON: 2001/11/25 ENTRY_REPORTED_BEGIN: The following code results in the second "Set" method appearing incorrectly as a standalone (global) function:
class Foo {
public:
void Set(int a);
void Set(char *cp, int a);
};
void Foo::Set(int a)
{
/** Set one. Red */
}
void Foo::Set(char *cp, int a)
{
/** Set two. Yellow */
}
I think the problem is in ccdoc::database::erase_from_path_map, where
the entire list of statements associated with a path map entry are
removed from the path map, when only the particular statement should
be.
My suggested fix is as follows:
void ccdoc::database::erase_from_path_map(ccdoc::statement::base* stmt)
{
if( stmt ) {
if( stmt->get_type() != statement::base::STMT_COMMENT_PKGDOC &&
stmt->get_type() != statement::base::STMT_COMMENT_PKGDOC_URL &&
stmt->get_type() != statement::base::STMT_COMMENT_PREFIX &&
stmt->get_type() != statement::base::STMT_COMMENT_SUFFIX ) {
// Issue 0041
// Ignore comments, they don't have meaningful names.
const char* pid = stmt->get_id();
if( *pid ) {
string id;
stmt->get_hier_id_no_pkgs(id);
if( id.size() ) {
path_map_itr_type itr = m_path_map.find(id);
if( itr != m_path_map.end() ) {
#ifdef CHRIS
statement::base::stmts_t &vec = (*itr).second;
statement::base::stmts_itr_t vitr = vec.begin();
for(;vitr != vec.end(); vitr++) {
if ((*vitr) == stmt) {
vec.erase(vitr);
break;
}
}
if (vec.size() == 0)
m_path_map.erase(itr);
#endif
}
}
}
statement::base::stmts_t& children = stmt->get_children();
statement::base::stmts_itr_t itr = children.begin();
for(;itr!=children.end();++itr) {
erase_from_path_map(*itr);
}
}
.
.
Incorporated the fix by Chris and added a verification test. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0118 ENTRY_TITLE: Resolve partially scoped references. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2001/11/26 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2001/11/26 ENTRY_REPORTED_BEGIN: This is similar to issue 0065.
Ccdoc does not manage partially scoped names as shown in the following example:
namespace test52
{
class A
{
public:
typedef int nested_type;
class nested_class {};
void fct();
};
class B
{
public:
/** Linked correctly */
A get_some_class();
/** Not linked correctly */
A::nested_type get_nested_type();
/** Not linked correctly */
A::nested_class get_nested_class();
/** Not linked correctly */
A::fct get_nested_fct();
};
}
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r26. Verified by test 52.
Added a heuristic that attempts to resolve partially scoped names. It works better than before but it is not perfect because ccdoc doesn't understand the using namespace clause. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0119 ENTRY_TITLE: Inconsistent command line syntax for -html paths under DOS. ENTRY_REPORTED_BY: Richard Cooper ENTRY_REPORTED_ON: 2001/11/27 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: In regards to issue 0115, replacing the backslashes with forward slashes in the -html switch does fix the problem. i.e. this works
But in DOS you can't use backslashes in the headers path. i.e. this doesn't workccdoc -db docs.db -html ./doc ..\Headers\*.h
This is not entirely satisfactory because now the two path descriptions in the command line need to be in different formats (the html path in UNIX form, the header path in DOS form). IMO the best fix would be to accept any paths in either format on any platform, but I have no idea how tricky that would be to implement.ccdoc -db docs.db -html ./doc ../Headers/*.h
Unfortunately this is somewhat difficult because ccdoc can't reliably determine whether it is running under cmd.exe or under a unix shell that is running on top of cmd.exe at run time.
If it is running under cmd.exe, it can safely replace backslashes with forward slashes but, if it is running under a unix shell that is running on top of cmd.exe it cannot.
I will look into whether ccdoc can be enhanced to allow you to specify UNIX style paths (with wildcards) under DOS. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0120 ENTRY_TITLE: Add support for the proposed todo tag in javadoc. ENTRY_REPORTED_BY: Gustavsson Mĺrten ENTRY_REPORTED_ON: 2001/11/28 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2002/11/11 ENTRY_REPORTED_BEGIN: I think that a todo tag would be very useful. It is a javadoc proposed tag: http://java.sun.com/j2se/javadoc/proposed-tags.html.
I use the same namespace between packages.
The namespace comment is valid only for the package in which it is defined. This is alright. The problem is that the documentation is a bit unclear about the scope for namespace doc comment reuse. I thought the namespace comment could be reused between packages.
It would also be nice if the template definition was shown just like a function declaration.
struct A {
/**
* Do stuff.
*/
void do_stuff();
};
struct B {
/**
* Do stuff.
* @deprecated See {@link A::do_stuff}.
*/
void fct();
};
This was actually reported by a number of folks.
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r27. Verified by test58.
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0124
ENTRY_TITLE: Scanner condition evaluates to a constant value.
ENTRY_REPORTED_BY: Michel Joly de Lotbiniere
ENTRY_REPORTED_ON: 2002/02/25
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2002/03/21
ENTRY_REPORTED_BEGIN:
Just compiled "right out of the box" ccdoc v0.8 r26 2001/11/28 with
IBM's VisualAge C++ 4.0 on Win32. There was only one message:
CPPC1104I:The condition evaluates to a constant value. ("phase1_scanner.cc", line 179.9: )
The code around that is:
// ================================================
// Eliminate the "\\\n" white space. This cannot
// be done in skip_ws() because of the backslash
// trigraph sequence "??/".
// ================================================
if( '\\' == ch ) {
char ch1 = get_char();
if( '\n' ) { <===== The warning is here
return get_token();
}
put_char(ch1);
}
Did you intend to write: if( '\n' == ch1 ) { etc.?
This was rolled in with the fixed for issue 129. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0125 ENTRY_TITLE: Hash mark displayed for default {@link...} reference. ENTRY_REPORTED_BY: Ethan Nagel ENTRY_REPORTED_ON: 2002/02/25 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2002/11/11 ENTRY_REPORTED_BEGIN: I noticed another minor issue that would be nice if it was cleared up. Inline @links seem to work pretty well (actually they work better than the documentation, but when using the "#" syntax to reference a member of the current class the default text leaves the "#" sign attached! I'm frankly not sure this is even really a bug and there are easy workarounds, but it is annoying. Perhaps I'm just using ccdoc wrong somehow. Here's an example:
class Sample
{
public:
/**
* Calls {@link #b}.
*/
int A();
/**
* Calls {@link #A A}.
*/
} ;
The documentation for A would have the "#" sign displayed, while the
documentation for B would give me the results I think you should get
automatically.
typedef struct
{
int foo1; //@- useful stuff 1
int foo2; //@- useful stuff 2
} MY_STRUCT;
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0127
ENTRY_TITLE: CcDoc does not recognize recognize template derivations.
ENTRY_REPORTED_BY: Joe Linoff
ENTRY_REPORTED_ON: 2002/03/01
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2002/11/11
ENTRY_REPORTED_BEGIN:
CcDoc does not recognize recognize template derivations.
The following example does not work properly:
template <class T>
struct A {
void fct1();
};
struct B {
void fct2();
};
struct C : public A<B>
{
void fct3();
};
You can work around this bug as follows:
But, as you can see, it is ugly. I will fix this in r27. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r27. Added test57 to verify it. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0128 ENTRY_TITLE: Order class entities the same way that were ordered in v07. ENTRY_REPORTED_BY: Luis Menina ENTRY_REPORTED_ON: 2002/03/01 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: May the class content be sorted not just alphabetically, but separating method and attributes ? I think it would be more clear and efficient (as it was in ccdoc v0.7).templatestruct A { void fct1(); }; struct B { void fct2(); }; struct C : public A /**@#-*/<B>/**@#+*/ { void fct3(); };
In DOS mode ccdoc used the "DIR /B /S /O:N" command to deduce wild card file names. I changed this to "DIR /B /O:N" to eliminate the spurious subdirectory searches. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0130 ENTRY_TITLE: Code section wraps in a strange way for small browser windows. ENTRY_REPORTED_BY: Doug Wieland ENTRY_REPORTED_ON: 2002/03/15 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2002/11/12 ENTRY_REPORTED_BEGIN: The ccdoc output for code segment can cause unreadable output if the browser window is too small.
Modify ccdoc to allow different comments for each item in a comma separated declaration. For example:
const string
/** This is one string */
string1 = "aaa",
/** This is another string */
string2 = "bbb";
Note that you can workaround this in the near term as follows:
// Workaround /** This is one string */ const string string1 = "aaa"; //** This is another string */ const string string2 = "bbb";ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0132 ENTRY_TITLE: support single-line prefix comments via //@+ ENTRY_REPORTED_BY: Esmond Pitt ENTRY_REPORTED_ON: 2002/09/24 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN:
Presently we have
string m_memberString; //@- the member documentation for this item
How about adding this:
//@+ the member documentation for this item
string m_memberString;
/** the member documentation for this item */
string m_memberString;
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0133
ENTRY_TITLE: Anonymous namespaces are not handled correctly.
ENTRY_REPORTED_BY: Esmond Pitt
ENTRY_REPORTED_ON: 2002/09/24
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2002/11/12
ENTRY_REPORTED_BEGIN:
We use anonymous namespaces extensively. Each source file has one or more of
these:
namespace
{
// local static stuff, classes, constants, ...
}
These are supposed to be merged within a single source file but distinct
across source files.
Problems which occur:
You need to do what the compiler does: auto-generate a namespace name based on the file name, so you can keep all the contents separate, and recognize and suppress an auto name when reporting (i.e. report each name as <anonymous>), something like:
<anonymous> namespace // documentation or 'undocumented'
// contents of this anonymous namespace
<anonymous> namespace // documentation or 'undocumented'
// contents of this anonymous namespace
<anonymous> namespace // documentation or 'undocumented'
// contents of this anonymous namespace
<anonymous> namespace // documentation or 'undocumented'
// contents of this anonymous namespace
This was an involved fix. The current implementation keeps all of the comments associated with a namespace and chains them in the documentation for that namespace. All of the members of the namespace are grouped (by package).
Two new switches were added: -[no]rptcfuns to provide users with some level of control over the namespace output. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0134 ENTRY_TITLE: Need a forward link to the contents section for classes and packages. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2002/11/11 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2002/11/12 ENTRY_REPORTED_BEGIN: When the class description is large, it is hard to find the contents section. The change will allow users to click to get to the table of contents. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r27. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0135 ENTRY_TITLE: Fix the copyright notice in the code. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2002/11/13 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2002/11/13 ENTRY_REPORTED_BEGIN: Recently I discovered a mail thread in the linux.debian.legal newsgroup discussing the ccdoc copyright notice. The author: "Branden Robinson", made some excellent points which I will directly quote here.
This license's intentions are clearly DFSG-free, but I think it leaves a couple of things unclear.I think that the answer is that the author doesn't really understand all of this legal stuff and really appreciates any help that he can get:-) ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r28. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0136 ENTRY_TITLE: Ccdoc gets confused by a nested struct. ENTRY_REPORTED_BY: Dennis Marsa ENTRY_REPORTED_ON: 2002/11/13 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2002/11/14 ENTRY_REPORTED_BEGIN: The following code fragment causes ccdoc problems:
- Permission to distribute UNmodified copies of the software is not granted. This is the only potential DFSG problem I can see.
- I'm not sure what is meant by "provided...that the distributor grants the recipient permission for further distribution as permitted by this notice". It almost sounds like a copyleft. If every distributor has to license the altered copies under these same terms, it is. If every distributor simply has to preserve the original copyright notice and license terms, but has permission to withhold permission for sub-licensees to make or distribute copies (modified or not), then it is not a copyleft. Either intention is DFSG-free.
In the event that the author does not intend to be using a copyleft, I suggest that he use the MIT/X11 license instead, which is short, sweet, and very close in form to the license he is using:
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE X CONSORTIUM BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.If the author does intend to be employing a copyleft, I wonder just for the sake of intellectual curiosity why he didn't use the GNU GPL.
class A {
protected:
A();
public:
virtual ~A() = 0;
private:
A(const A& rhs);
A& operator=(const A& rhs);
public:
struct B {
int type;
};
};
The problems are:
namespace testXX {
class A {
public:
A();
~A();
/**
* Run the test using the specified arguments.
* @param argc The number of arguments.
* @param argv The collection of arguments.
* @returns The run status.
*/
virtual int Run(int argc,char** argv);
};
class test : public A {
public:
test();
~test();
/**
* Run the test using the {@link test99::A::Run} command.
* @param argc The number of arguments {@link test99::A::Run}.
* @param argv The collection of arguments {@link test99::A::Run}.
* @returns The run status {@link test99::A::Run}.
*/
virtual int Run(int argc,char** argv);
};
};
namespace testXX {
class A {
public:
A();
~A();
/**
* Run the test using the specified arguments.
* @param argc The number of arguments.
* @param argv The collection of arguments.
* @returns The run status.
*/
virtual int Run(int argc,char** argv);
};
class test : public A {
public:
test();
~test();
/**
* Run the test using the {@link test99::A::Run} command.
* @param argc The number of arguments
* {@link test99::A::Run}.
* @param argv The collection of arguments
* {@link test99::A::Run}.
* @returns The run status {@link test99::A::Run}.
*/
virtual int Run(int argc,char** argv);
};
};
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0138
ENTRY_TITLE: The @see and {@link} directives do not correctly reference entities in different packages.
ENTRY_REPORTED_BY: Markus Stulle
ENTRY_REPORTED_ON: 2003/01/09
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2003/01/13
ENTRY_REPORTED_BEGIN:
We are facing the following problem while documenting
source code produced using programming language ansi-c: if a function
'foo()' is documented in a named packaged, e.g. 'A.B.C', we fail in
referencing that function using the '@link' directive:
// file 'foo.c': /** * @pkg A.B.C */ /** * this function does nothing, * so it is error-free by design. */ void foo( void );
// file 'bar.c':
/**
* @pkg D.E
*/
/**
* again an important procedure.
* functionality is identical to {@link A.B.C#foo foo}.
*/
void bar( void );
We tried the following entity strings for referencing 'foo', but none of
them produced a link in the resulting hypertext document:
The motivation for our investigations is the fact, that, unfortunaltely, 'foo()' is defined multiply in our project (i.e. in different packages) and we want to have a direct link to it from 'bar()' generated by the 'ccdoc' processor. if we employ the directive '{@link #foo foo}' ccdoc works fine but delivers a link to a page where the user can select from one of the multiply defined foo()s. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r29 and verified by test 65.0) A.B.C.foo 1) A.B.C#foo 2) #A.B.C.foo 3) A::B::C#foo 4) A::B::C.foo 5) A::B::C::foo 6) #A::B::C::foo
The problem was that these references were not considering the path info when looking up a name in the ccdoc::database::get_stmt_no_pkgs() method. This method was updated to try harder when a name was not found by looking up the name using the path information. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0139 ENTRY_TITLE: Core dump when an invalid package name of the form "A::B:" appears. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/01/13 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/01/13 ENTRY_REPORTED_BEGIN: ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r29 and verified by test 66.
There was a missing break that caused the iterator in void ccdoc::database::parse_path() to run off of the end. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0140 ENTRY_TITLE: Core dump when a friend declaration of the form: foo&operator<< appears. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/01/14 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/01/14 ENTRY_REPORTED_BEGIN: ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r29 and verified manually. This caused the release date to bump up to 2003/01/15.
There was a missing break that caused the iterator in void ccdoc::database::parse_path() to run off of the end for angle brackets. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0141 ENTRY_TITLE: Architecture is not reflected in the bin directory name. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/01/21 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/01/21 ENTRY_REPORTED_BEGIN: The problem is that I can't differentiate between linux on an i386 platform, linux on MacOS X and linux on a sparc station. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed by using $Config{'archname'} instead of $Config{'osname'} in the platform script. This change doesn't affect users to it will be released as part of r30 when the StrongARM port is done. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0142 ENTRY_TITLE: No man page for linux ports. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/01/21 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/01/21 ENTRY_REPORTED_BEGIN: The Debian linux port had to create its own man page for ccdoc. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: I added a man page that is automatically derived from the help text. This change doesn't affect users to it will be released as part of r30 when the StrongARM port is done. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0143 ENTRY_TITLE: Support javadocs frames format. ENTRY_REPORTED_BY: Thomas Bednarz ENTRY_REPORTED_ON: 2003/01/24 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: "Javadoc documentation is great and very handy to use".
Since static indicates class scope for attributes and methods on classes it is in a category of its own. Class scoped attributes are now marked in the contents table. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0145 ENTRY_TITLE: It is impossible to reference an operator in the @see and {@link..} directives. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/01/30 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/01/30 ENTRY_REPORTED_BEGIN: I cannot reference an operator== in the @see and {@link..} directives. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r31 and verified by test 67.
The problem was that the operator keyword was not understood by the lookup ccdoc::database::get_stmt_no_pkgs() method. In ccdoc, all operator names consist of two parts: the operator keyword and the operator tokens. They are separated by a space. The link syntax for @see and {@link} do not allow spaces so special dispensation was made to allow users to specify links without spaces. The system then figures them out during the stmt lookup. This example below shows how to specify links for operators:
namespace test67
{
class Test
{
public:
/**
* Is equal, this is the same as {@link #is_equal is_equal}.
* @see #is_equal
*/
bool operator==(const Test&) const;
/**
* Test method, this is the same as {@link #operator== ==}.
* @see #operator==
*/
bool is_equal(const Test&) const;
};
}
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0146
ENTRY_TITLE: Namespace processing doesn't work properly in certain cases.
ENTRY_REPORTED_BY: Sakari Forsberg
ENTRY_REPORTED_ON: 2003/02/07
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2003/02/10
ENTRY_REPORTED_BEGIN:
The current version of ccdoc fails to correctly associated entities
between two entities in the same namespace in the following example:
// file1.h
namespace A {
int B;
}
// file2.h
namespace A {
/**
* @see #B
*/
int C;
}
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r32 and verified by test 68.
The problem was a bug processing undocumented namespaces. In older versions it can be worked around by specifying the -rptcfuns switch or by adding a comment to one of the namespaces. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0147 ENTRY_TITLE: Lots of compiler warnings when building ccdoc using the Borland C++ compiler. ENTRY_REPORTED_BY: Sakari Forsberg ENTRY_REPORTED_ON: 2003/02/07 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/10 ENTRY_REPORTED_BEGIN: Saw lots of warnings when compiling ccdoc using the Borland C++ compiler. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r32.
This was occurring because I explicity turned off warning messages in the compilers that I use (g++,sun CC,Mac OSC c++) for some silly reason. I turned on warnings and cleaned up all of the warning messages for this release. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0148 ENTRY_TITLE: Allow users to customize ccdoc output using a CSS file. ENTRY_REPORTED_BY: Sakari Forsberg ENTRY_REPORTED_ON: 2003/02/10 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: Have you thought, instead of giving color parameters etc. to ccdoc that it can make custom looking html you could give a custom tailored .css-file to ccdoc and it could generate html which uses a .css-file?
template <class T>
class Test
{
public:
Test();
~Test();
T GetValue() const {return m_val;}
void SetValue(T val);
private:
T m_val;
};
template <class T>
Test<T>::Test()
{
}
template <class T>
Test<T>::~Test()
{
}
template <class T>
void Test<T>::SetValue(T val)
{
m_val = val;
}
template <class T>
T real_fct(T val)
{
Test<T> f;
f.SetValue(val);
return f.GetValue();
}
template<> int real_fct<int>(int);
template<> class Test<double>;
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r33 and verified by test 69.
I added a phase 1 switch called -[no]tcms to tell ccdoc to ignore template class methods that are implemented outside of the declaration. The default for this switch is -notcms.
This implies that all template method documentation must be done in the declaration. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0150 ENTRY_TITLE: Make it easier to update the revision number in the source files. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/02/17 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/18 ENTRY_REPORTED_BEGIN: During the last few releases I noticed that updating the various source files with the new revision numbers is error prone. To make it less error prone, I would like to automate it. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r34.
Added the release.pl script to update various source files with the new revision number. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0151 ENTRY_TITLE: Translate the generated html to other languages. ENTRY_REPORTED_BY: Sakari Forsberg ENTRY_REPORTED_ON: 2003/02/14 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: Have you thought adding a change to translate contents of generated html? It would be nice, if ccdoc would read certain words like "class", "contents" etc. from a file which would be kind of resource file. If you change contents a file to f. ex. finnish, then you would have finnish documentation!
The file (simple ex.) could have key-string pairs:
Then you could generate english documentation: ccdoc -langfile ccdoc.en.language ... and finnish: ccdoc -langfile ccdoc.fi.language ... ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0152 ENTRY_TITLE: Ignore duplicate macros names. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/02/17 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/26 ENTRY_REPORTED_BEGIN: This occurs when the -rptmac switch is specified and windows DLL import/export macros are defined in header files as import by default without ccdoc guards. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Originally fixed in r34 (2003/02/18) and verified by test 71.file: ccdoc.en.language CLASS=class DESTRUCTOR=destructor file: ccdoc.fi.language CLASS=luokka DESTRUCTOR=tuhoaja
Fixed again in r39 and verified by test 86 because I found another problem related to this when the same macro was defined in different packages. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0153 ENTRY_TITLE: Add a heuristic to ignore include guards and DLLIMPORT macros. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/02/17 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/18 ENTRY_REPORTED_BEGIN: Add a heuristic that ignores certain macros under switch control. Documentation for include guards and windows DLLIMPORT/DLLEXPORT macros do not help large systems.
This heuristic would ignore macros named with the following rules:
In the following example only the DBG macro would be documented:
I propose that the switch be called -rptmac1 (report macros heuristically) in phase 1. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r34 and verified by test 72.#ifndef file_h #define file_h #ifdef WINDOWS #include# ifndef file_dll # define file_dll DLLIMPORT # else # define file_dll DLLEXPORT # endif #endif /** * This macro defines the debug prefix. */ #define DBG cerr << "DEBUG:" << __FILE__ << ":" << __LINE__ << " " #endif
This is my best guess at reasonable macro names. In the future it might be better to make this file driven and use these as the default if the file is not specified. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0154 ENTRY_TITLE: Eliminate the requirement that the -rptmac and -rptmac1 switches appear in phase 1. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/02/18 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/19 ENTRY_REPORTED_BEGIN: Requiring that the -rptmac and -rptmac1 switches appear in phase 1 and phase 3 processing is inconsistent and could lead to user confusion and unexpected results. These switches should only be required in phase 3. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r35. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0155 ENTRY_TITLE: Incorrect documentation when there is only one ccdoc namespace comment for namespace declarations in multiple files. ENTRY_REPORTED_BY: Sakari Forsberg ENTRY_REPORTED_ON: 2003/02/18 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/19 ENTRY_REPORTED_BEGIN: When a namespace is defined in more than one header file but is only documented in one place, ccdoc auto generates comments and puts the user documentation in a table. The test case below illustrates the problem when the ccdoc is generated.
-----------------------
// file1.h
/**
* This is the namespace documentation.
* @author Author
* @version 1.0
*/
namespace Test {
int fct1();
}
-----------------------
// file2.h
// No ccdoc documentation here. It should derive
// the documentation from file1.h
namespace Test {
int fct2();
}
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r35, verified by test 75.
ENTRY_RESOLVED_END:
# ================================================
ENTRY_ID: 0156
ENTRY_TITLE: More namespace problems in r35.
ENTRY_REPORTED_BY: Sakari Forsberg
ENTRY_REPORTED_ON: 2003/02/19
ENTRY_STATUS: FIXED
ENTRY_RESOLVED_BY: Joe Linoff
ENTRY_RESOLVED_ON: 2003/02/19
ENTRY_REPORTED_BEGIN:
Namespace processing gets confused in certain cases.
namespace N1 {
class Class1
{
public:
Class1();
~Class1();
};
}
namespace N2 {
class Class2 : public N1::Class1
{
public:
Class2();
~Class2();
};
}
/**
* This namespace should contain Class1.
* @author Author
* @version 1.0
*/
namespace N1 {
class Class1
{
public:
Class1();
~Class1();
};
}
/**
* This namespace should contain Class2
* with a reference to N1::Class1.
* @author Author
* @version 1.0
*/
namespace N2 {
class Class2 : public N1::Class1
{
public:
Class2();
~Class2();
};
}
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r36, verified by test 76.
This was a critical bug that warrants a new release. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0157 ENTRY_TITLE: In some cases operator code is being processed incorrectly. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/02/19 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/19 ENTRY_REPORTED_BEGIN: In some cases the body of an operator is parsed incorrectly because there was a portion of code in the parser that exited prematurely before processing the body.
This has been fixed and will be released in r37 later this week. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r37, verified by test 69. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0158 ENTRY_TITLE: Class method implementations are reported as global functions. ENTRY_REPORTED_BY: Richard Durwin ENTRY_REPORTED_ON: 2003/02/19 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/19 ENTRY_REPORTED_BEGIN: In the following code example ccdoc reports the class method implementations as global functions.
class Test {
public:
Test();
Test(const Test& obj);
~Test();
Test& operator=(const Test& obj);
void Set(int v);
int Get() const;
private:
int m_a;
};
namespace {
const char* rcsid = "$Id: issues.txt,v 1.34 2004/09/30 16:09:26 jlinoff Exp $";
};
Test::Test()
: m_a(0)
{
}
Test::Test(const Test& obj)
: m_a(obj.m_a)
{
}
Test::~Test()
{
}
Test& Test::operator=(const Test& obj)
{
m_a = obj.m_a;
return *this;
}
void Test::Set(int v)
{
m_a = v;
}
int Test::Get() const
{
return m_a;
}
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r37, verified by test 77.
Modified the parser to ignore these. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0159 ENTRY_TITLE: Very long names make the contents table very hard to read. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/02/20 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/20 ENTRY_REPORTED_BEGIN: When ids are very long it makes the contents table very difficult to read. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r37, verified by tests 78,79,80.
Added a switch called -rptmlcei <num> (maximum length of the contents entity id) that limits the size of the entity id. When the string exceeds that length, only the first <num> characters are printed followed by .. to indicate that it has been truncated. The default length is 32. If no inherited from column exists, the value of the -rptmlcifi is added to make this field bigger.
Also added a switch called -rptmlcifi <num> (maximum length of the contents inherited from id) that limits the size of the inherited from column. When the string exceeds that length, only the first <num> characters are printed followed by .. to indicate that it has been truncated. The default length is 32. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0160 ENTRY_TITLE: Empty function names appear under certain conditions. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/02/20 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/20 ENTRY_REPORTED_BEGIN: Empty function names appear when the following syntax is encountered:
#ifdef FOO
int Fct(int arg)
#else
long Fct(long arg)
#endif
{
return arg;
}
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r37, verified by test 81.
This occurs because ccdoc keeps around the pre-processing directives. The current solution is to print a warning and let the user fix the problem. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0161 ENTRY_TITLE: Strange function names like "max)" appear sometimes. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/02/20 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/20 ENTRY_REPORTED_BEGIN: While parsing some STL header files I came across the following construct which caused ccdoc to become confused about the method name. It reported the name as "max)" instead of "max".
class Test {
public:
static double (_STLP_CALL max)() _STLP_NOTHROW { _STLP_USING_VENDOR_CSTD return DBL_MAX; }
};
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r37, verified by test 81.
This was occurring because the get_fct_name method in phase 1 parsing was getting confused under certain conditions. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0162 ENTRY_TITLE: Doxygen compatibility - support single line comments using doxygen syntax. ENTRY_REPORTED_BY: Richard Durwim ENTRY_REPORTED_ON: 2003/02/21 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/21 ENTRY_REPORTED_BEGIN: Single line comments use a different flag. That's a lot of edits to make!
Here are the flavors that ccdoc will support:
ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r37, verified by test 82. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0163 ENTRY_TITLE: Doxygen compatibility - ignore the @endlink directive. ENTRY_REPORTED_BY: Richard Durwim ENTRY_REPORTED_ON: 2003/02/21 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/21 ENTRY_REPORTED_BEGIN: @link uses @endlink to delimit it, rather than the end-of-line. The @endlink appears in the ccdoc output.int var; /*!< Detailed description after the member */ int var; /**< Detailed description after the member */ int var; //!< Brief description after the member int var; ///< Brief description after the member
Modified ccdoc to ignore the @endlink directive.
In r38 updated ccdoc to officially recognize the @endlink directive as part of normal processing. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0164 ENTRY_TITLE: Doxygen compatibility - ignore @file comment blocks. ENTRY_REPORTED_BY: Richard Durwim ENTRY_REPORTED_ON: 2003/02/21 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/21 ENTRY_REPORTED_BEGIN: Doxygen uses @file to document files. You could just ignore this flag. At the moment it appears in ccdoc documentation.
Modified ccdoc to ignore the @file comment blocks. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0165 ENTRY_TITLE: Doxygen compatibility - support virtual grouping of definitions. ENTRY_REPORTED_BY: Richard Durwim ENTRY_REPORTED_ON: 2003/02/21 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: Doxygen allows a virtual grouping of definitions that is very useful (@name and @{).
static const char* s_rcsid = {"value"};
The resulting variable declaration strips out the {"value"} so it looks
like this:
The good news is that the variable is correctly identified and the comments are correctly associated. The value is the only thing missing. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r38, verified by tests 2 and 83.static const char* s_rcs_id =
Ccdoc was discarding what it thought was the function body for this special case. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0167 ENTRY_TITLE: In German we want to use embedded dots in the short description. ENTRY_REPORTED_BY: Christian Ehrlicher ENTRY_REPORTED_ON: 2003/02/25 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: I've found: another little problem. In German we write for the short form of "for example" "z.B.". ccdoc ignores silently the rest of the line after the B:
--> in .h : /** das ist z.B. test */
--> in .html : "das ist z.B" without the last point an the rest of the line
For now you can workaround it by embedding trailing spaces before the last dot as shown below:
--> in .h : /** das ist z.B . test */
--> in .html : "das ist z.B. test"
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r39, verified by test 84.
The short description string size was not updated properly in ccdoc::phase1::scanner_doc::add_line. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0168 ENTRY_TITLE: Slow g++ 3.2 compile of help.cc because of the large help string. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/02/25 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/25 ENTRY_REPORTED_BEGIN: Very slow compilation of help.cc using g++ 3.2 because it tries to verify the format for the very large help string. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r39.
Modified the script that generates the help text in help.cc to break it up periodically. This improved compile times by more than 100X on my system. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0169 ENTRY_TITLE: Man page has formatting problems. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/02/25 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/25 ENTRY_REPORTED_BEGIN: The automatically generated man page has formatting problems. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r39.
Modified the script that generates the man page to recognize the constructs that were missed in the previous version. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0170 ENTRY_TITLE: Bug on NT platforms parsing command line with spaces and wildcards. ENTRY_REPORTED_BY: Gonzalo Matamala Torres ENTRY_REPORTED_ON: 2003/02/26 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/26 ENTRY_REPORTED_BEGIN: Hello (and sorry by my poor english)!
I have found a bug over NT platforms parsing command line with spaces and wildcards:
Unfortunately, the code I'm documenting has white space in paths. (It's a horrible name coding, I know it!). But I need document it. For example, library code in:
smartcard components\scot
when I try, for example:
ccdoc try resolve wildcard with (switches.cc, line 411):ccdoc -db example.db "smartcard components\scot\*.h"
that result (in my example):cmd = "DIR /B /O:N " + file + " > " + tmp + "\n";
and dir command try list "smartcard" directory first and "components\scot\*.h" after Then, fail because these paths aren't found.DIR /B /O:N smartcard components\scot\*.h
A solution is using -files switch of course. But I think the above line can be modified to:
that result:cmd = "DIR /B /O:N \"" + file + "\" > " + tmp + "\n";
(double quotes are added to path) and then it's a success.DIR /B /O:N "smartcard components\scot\*.h"
Thank you for taking the time to thoroughly analyze the problem and propose a solution. I will add this fix to the v08r39 release so it should be available to you next week. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r39.
Manually tested. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0171 ENTRY_TITLE: Constructors with constructed default arguments are reported with the wrong method name. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/02/26 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/02/26 ENTRY_REPORTED_BEGIN: Constructors with constructed default arguments are reported with the wrong method name as shown in the example below:
class Test
{
public:
typedef Allocator<Property*> allocator_type;
typedef Allocator<Association*> allocator_type1;
Test( const allocator_type& x = Allocator<Property*>() ) : m_val(0) {}
Test( const allocator_type1& x = allocator_type1() ) : m_val(0) {}
private:
int m_val;
};
ENTRY_REPORTED_END:
ENTRY_RESOLVED_BEGIN:
Fixed in r39, verified by test 85.
Modified ccdoc::phase1::parser::get_fct_id to correctly recognize the constructor situation. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0172 ENTRY_TITLE: Template friend declarations are not resolved properly when packages are used. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/03/11 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/03/11 ENTRY_REPORTED_BEGIN: Template friends are not resolved properly when packages are used. They work correctly when packages are not used. Here is some source code that demonstrates the problem:
template <class T>
class A {
friend class B<T>;
T m_val;
};
template <class T>
class B {
T m_val1;
A<T> m_val2;
};
Here is the ccdoc command that triggers the problem:
If the -pkg switch is removed, the reference is found correctly. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r40, verified by test87.% ccdoc -db foo.db *.h -pkg test -index -html html/
The phase1 parser was not correctly stripping the names to eliminate template arguments. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0173 ENTRY_TITLE: The md5hash macro is being improperly ignored by the rptmac1 heuristic. ENTRY_REPORTED_BY: Alexander Filonov ENTRY_REPORTED_ON: 2003/04/09 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/05/15 ENTRY_REPORTED_BEGIN: I've encountered a problem with ccdoc using -rtpmac1 heuristics:
Macro called "md5hash" will not appear in final html code.
I see two solutions for this:
The backwards comparison heuristic was not working properly. The option for telling ccdoc to ignore inline text commands already exists. You can use the predefined __ccdoc__ macro for this as follows:
ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0174 ENTRY_TITLE: Bug#195282: ccdoc: FTBFS with g++-3.3 ENTRY_REPORTED_BY: Magnus Ekdahl ENTRY_REPORTED_ON: 2003/05/30 ENTRY_STATUS: FIXED ENTRY_RESOLVED_BY: Joe Linoff ENTRY_RESOLVED_ON: 2003/06/05 ENTRY_REPORTED_BEGIN:#ifndef _include_guard # ifndef __ccdoc__ # define _include_guard # endif #endif
Package: ccdoc Version: 0.8.38-1 Severity: serious >From my build log: ... source='statement.cc' object='statement.o' libtool=no \ depfile='.deps/statement.Po' tmpdepfile='.deps/statement.TPo' \ depmode=gcc3 /bin/sh ./depcomp \ g++ -DHAVE_CONFIG_H -I. -I. -I. -DCCDOC_CID=\"/usr/bin\" -g -O2 -c -o statement.o `test -f 'statement.cc' || echo './'`statement.cc statement.cc: In static member function `static bool ccdoc::statement::base::is_rptmac1_id(const char*)': statement.cc:987: error: brace-enclosed initializer used to initialize `const char*' statement.cc:987: error: brace-enclosed initializer used to initialize `const char*' ... make[2]: *** [statement.o] Error 1 make[2]: Leaving directory `/tmp/buildd/ccdoc-0.8.38/src' make[1]: *** [all] Error 2 make[1]: Leaving directory `/tmp/buildd/ccdoc-0.8.38/src' make: *** [build-stamp] Error 2 -- System Information: Debian Release: testing/unstable Architecture: i386 Kernel: Linux frobnitz 2.4.21-pre5 #1 Sat Mar 1 09:01:10 PST 2003 i686 Locale: LANG=C, LC_CTYPE=CENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: Fixed in r41.
The array initialization was incorrect (in r40), g++ 3.3 was correct in reporting it. ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0175 ENTRY_TITLE: Generate an index in HTML and XML. ENTRY_REPORTED_BY: Joe Linoff ENTRY_REPORTED_ON: 2003/06/05 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: Several users have requested a search capability in ccdoc. This capability can be provided in two ways: use a standard HTML based search engine on the generated files or provide a database of indexed items that can be used for a custom search engine. Although the standard HTML based search engine approach works well, there are still some users who would like a pre-generated index of searchable items.
My current thinking about how to implement this is to provide a -rptidx switch for phase 3 that generates two HTML indices: one by name and one by entity type by name.
It would also generate an XML based index file that could be used by 3rd party search engines. Each XML record would contain the name of the entry (i.e. GetName), the type (method, function, etc.) and the associated URL. Multiple entries with the same name would be allowed.
It might also be desirable to have it contain information about the
scope of the name. This means that each enclosing namespace and class
might be specified as list. Consider the case of the X::Y::A::B::GetName()
method where X and Y are namespaces, and B is a nested class inside of
A. The scope information would contain data like
((nsp X) (nsp Y) (cls A) (cls B)).
Yet another desirable attribute would be to have this index view of the data tie in to the XML output that is being generated in the proto-type version for the next major release. In this release the database format will convert to XML. The phase 2 processing will generate fully cross-indexed entities that are used in phase 3 to generate HTML. Of course the user can stop at phase 2 and work directly with the XML. This is different than today where the database format is NOT XML and it is not fully cross referenced. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0176 ENTRY_TITLE: Enhance the table of contents for easier navigation with letter based links at the top. ENTRY_REPORTED_BY: Ragav Satish ENTRY_REPORTED_ON: 2003/05/23 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: This suggestion was actually made some time ago but I just got around to adding it.
Ragav suggested that ccdoc be enhanced to support quick links to entries in the table of contents. The idea is that for tables with a large number of classes (or methods), the user could click on a quick link to the region they were interested in.
In practice it would look something like this:
I think that this would be very useful for large systems that have hundreds or even thousands of classes. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0177 ENTRY_TITLE: Invalid link from parameter variable name to member variable. ENTRY_REPORTED_BY: Chris Rogers ENTRY_REPORTED_ON: 2003/06/18 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: I have found something wrong with my output though - not sure if this is a bug or if I am doing something wrong. I have the following (simplified) situation:Table Of Contents Quick Links: A B C D E F G H ... Z . . .
class SOMECLASS
{
public:
void DoSomething(int x);
int x;
};
You'll see that I have a member variable 'x' and also a member
function which has 'x' in it's argument list. This causes CcDoc to
place a link from the argument name 'x' to the documentation for the
member variable 'x'.
I could simply change one of these names to ensure that there is no clash, however I thought you'd like this kind of problem submitted as a bug. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0178 ENTRY_TITLE: Namespace/anonymous enum interaction bug. ENTRY_REPORTED_BY: William A. Rivet ENTRY_REPORTED_ON: 2003/06/26 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: I have a file as_config.h:
#ifndef AS_CONFIG_H
#define AS_CONFIG_H
namespace ApplicationInterface
{
enum
{ zero=0,
one=1,
two=2
};
enum {
number_of_tasks = 60,
task_stack_size = 4096,
number_of_queues = 32,
number_of_event_groups = 32,
number_of_timers = 32
};
} // namespace ApplicationInterface
#endif
The problem is that I see two "$anonymous$" enum entities in the namespace page, but both link only to a page that lists only the first enum.
I know that in this case the enums could be combined, but this seems less than satisfactory as a general solution. Is this a known issue? (I couldn't find one that seemed to match on the ccdoc website) Is there interest in fixing this? I have no idea how hard it is to change this. Either way I likely will have to do something to at least address this in-house.(perhaps just complain when the second enum is detected...this would at least allow us to detect the problem and perhaps "invent" a name for at least one of them). ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0179 ENTRY_TITLE: Add a new @entity directive to allow referencing. ENTRY_REPORTED_BY: Steve Perry ENTRY_REPORTED_ON: 2003/07/15 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: It would be nice if ccdoc had a directive like "@entity <entityName>".
Here is how I would want to use it. I have a macro which declares and defines a static member function. It is a macro because every class in the hierarchy should have this function (it returns an "interface ID"). The code currently looks like this:
//@{
// Returns the ID value for this interface.
//@}
Gf_DECL_INTERFACE_ID("{8DC3F1C6-F7CA-42d8-AFD6-7AF950C5C11E}")
# if defined(__ccdoc__)
static Gf_GUID InterfaceID();
# endif
It would be nicer if I could write:
//@{
// Returns the ID value for this interface.
// @entity static Gf_GUID InterfaceID()
//@}
Gf_DECL_INTERFACE_ID("{8DC3F1C6-F7CA-42d8-AFD6-7AF950C5C11E}")
If there is already a way to do this, I'd appreciate hearing about it. If not, please consider adding a feature of this sort sometime in the future.
I've tried using the -maxpathlen option during the (-htm) output generation phase, but this doesn't shorten the filename length. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0181 ENTRY_TITLE: Copy constructor documented incorrectly under some conditions. ENTRY_REPORTED_BY: Charles Doland ENTRY_REPORTED_ON: 2003/08/15 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: I would like to report a minor issue with ccdoc, described below.
The documents created by ccdoc indicate that a constructor will be generated automatically by the compiler even if one is specified. This seems to occur when the constructor for class X is specified as X(X const &), but does not occur when the constructor is specifed as X(const X &), although these are equivalent. ENTRY_REPORTED_END: ENTRY_RESOLVED_BEGIN: ENTRY_RESOLVED_END: # ================================================ ENTRY_ID: 0182 ENTRY_TITLE: Add directory tree/outline expansion (+/-) capabilities to better organize the information. ENTRY_REPORTED_BY: Brian Alexander ENTRY_REPORTED_ON: 2004/06/04 ENTRY_STATUS: OPEN ENTRY_RESOLVED_BY: ENTRY_RESOLVED_ON: ENTRY_REPORTED_BEGIN: I was going through the ccdoc for [one of my classes] this morning when I had a couple of feature ideas for ccdoc. I was thinking that it would be nice if overloaded methods in a class could be collapsed and expanded so that you only saw one line with the function name. If all of the short descriptions matched for the method, you could see the short description - if the short descriptions didn't match you would see something like "6 overloads". E.g. you might see something like this:
When you click the + to expand you would see the display much like it is today with each overload displayed on its line with its short description. At this point the '+' would turn into a '-' and clicking the '-' would collapse it back down.+ GetClient Xdm_ObjectDerived method protected 16 overloaded methods (click '+' to expand)
Extending that idea, I thought it would be nice if we had an extra tag that we could tag methods with to categorize them - E.g. "@category Debug Helper". All methods with matching categories would go into an expandable / collapsable grouping like the overloaded methods. (If there's just one or two methods in a category maybe it defaults to expanded but more than two and the category defaults to collapsed.) Methods with no category would be displayed as they are today without the expand/collapse capability. Methods within a category are alphabetized within the category. We would also probably want links at the top and bottom to 'expand all' and 'collapse all' - maybe throw in some cookies to store the user's preferences (overall and for each class) - hah.
Then I also thought that it was somewhat annoying with the ccdoc for some classes - Xdm_Model is one example - where there is a lot of good documentation for the class but it's a pain to have to scroll through all of that to try and find the method index so I can see if it supports the method I'm looking for. Having all of the class documentation and examples are great for the first time you're using a class but for quick reference which is most of the time after the first time you've used a class, all of that stuff gets in the way. If classes could have extra tags to separate things out like @Examples, @Extending, @UserGuide or something like that then in the ccdoc, you could either have expand / collapse things like above to expand out the examples text - or you could have links that take you to the examples for a specific class etc. I like the expand/collapse symbols just to keep everything on one page but it would take some javascript or something I'm sure and may be more pain than it's worth.
Anyway, these were just some pie in the sky ideas I had this morning. I didn't think them through a whole lot - just wanted to send them off to you to see if you liked them.
There was a particular order of entities in the class that was causing consternation in the compare method used during the sorting of the table of contents in phase3_html.c. I was able to duplicate the problem (test88) and then fix it by using a std::stable_sort() instead of a std::sort(). It is interesting to note that this failure occurred on solaris and linux but it did not occur on windows. ENTRY_RESOLVED_END: