summaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* refactoring: modernize search indexDimitri van Heesch2020-10-032-157/+116
| | | | Also QVector has been removed from qtools as it was no longer used.
* Merge branch 'master' of github.com:doxygen/doxygenDimitri van Heesch2020-10-022-2/+2
|\
| * Merge pull request #8068 from albert-github/feature/bug_loss_dataDimitri van Heesch2020-10-011-1/+1
| |\ | | | | | | Warning about possible loss of data
| | * Warning about possible loss of dataalbert-github2020-10-011-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | Om 64-bit windows platform we get the warning: ``` src\template.cpp(341): warning C4267: 'return': conversion from 'size_t' to 'uint', possible loss of data ``` Explicit setting conversion..
| * | Warning in internal documentationalbert-github2020-10-011-1/+1
| |/ | | | | | | | | | | | | | | | | | | | | | | The internal documentation gives warning: ``` .../src/clangparser.h:33: warning: argument 'fileName' of command @param is not found in the argument list of ClangTUParser::switchToFile(FileDef *fd) .../src/clangparser.h:35: warning: The following parameter of ClangTUParser::switchToFile(FileDef *fd) is not documented: parameter 'fd' ``` this is corrected. (Note: warning regarding TokenManager is an javaCC upstream problem and corrected in their master).
* | issue #7706: Md5 hash does not match for two different runsDimitri van Heesch2020-10-021-2/+2
|/
* issue #8066: Doxygen crashes in ClangTUParserDimitri van Heesch2020-09-302-4/+10
|
* Merge pull request #8060 from albert-github/feature/bug_md_open_bracketDimitri van Heesch2020-09-291-0/+2
|\ | | | | No warning for unbalanced brackets in markdown
| * No warning for unbalanced brackets in markdownalbert-github2020-09-251-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When we have the input files qq.md. ``` @page xxx0 yyy0 @} Start of text @page xxx1 yyy1 @{ Start of text ``` and qqh.h: ``` /** @page xxxh0 yyyh0 @} Start of texth */ /** @page xxxh1 yyyh1 @{ Start of texth */ ``` we get the warnings: ``` qq.md:3: warning: unbalanced grouping commands qqh.h:4: warning: unbalanced grouping commands qqh.h:15: warning: end of file with unbalanced grouping commands ``` ``` so we are missing ``` qq.md:9: warning: end of file with unbalanced grouping commands ``` due to the fact that the closing routine was not called (also the open routine was not called, always good to call it the set variables to their proper values.
* | Incorrect sorting of reflist itemsalbert-github2020-09-291-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When having an example like: ``` /// \file /** \xrefitem my_errors "err2" "ERR1" ERROR 101*/ #define MY_ERR_CANNOT_OPEN_FILE 101 /** \xrefitem my_errors "err2" "ERR2" ERROR 102*/ #define MY_ERR_CANNOT_CLOSE_FILE 102 /** \xrefitem my_errors "err2" "ERR1a" ERROR 101a*/ #define MY_ERR_CANNOT_OPEN_FILE3 103 /** \xrefitem my_errors "err2" "ERR2a" ERROR 102a*/ #define MY_ERR_CANNOT_CLOSE_FILE4 104 ``` the sorting should be done on the part `MY_ERR_CANNOT_...` but no sorting is done (since 1.8.18) as: - call to `qstricmp` contains twice the same value - a boolean value should be the result of the compare but in fact an integer value was returned This problem is most likely a regression due to ``` Commit: aca13723a9373a1080ca7f108e7be0905b9ae793 [aca1372] Date: Thursday, February 27, 2020 10:38:22 PM Restructure the way RefLists are handled ```
* | issue #8063: Failed to build current "master" branchDimitri van Heesch2020-09-283-13/+19
| |
* | Fix for broken LaTeX output.Dimitri van Heesch2020-09-2720-42/+42
| | | | | | | | | | | | Fixes error: "\begin{DoxyCodeInclude} on input line 93 ended by \end{DoxyCode}" while generating the LaTeX version of the manual
* | Merge branch 'multithread_outputgen'Dimitri van Heesch2020-09-2734-750/+862
|\ \ | |/ |/|
| * Refactoring: prepare output generators for multi-threaded useDimitri van Heesch2020-09-2734-750/+862
| |
* | Merge pull request #8045 from albert-github/feature/bug_md_fenced_cntDimitri van Heesch2020-09-241-1/+1
|\ \ | | | | | | Miscounting of lines in case of fenced code block
| * | Miscounting of lines in case of fenced code blockalbert-github2020-09-201-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When we have an example like: ~~~ Page ======= test text test text \error1 ``` make ``` ``` make ``` test text \error2 ~~~ we get the warnings: ``` aa.md:5: warning: Found unknown command '\error1' aa.md:17: warning: Found unknown command '\error2' ``` instead of ``` aa.md:5: warning: Found unknown command '\error1' aa.md:15: warning: Found unknown command '\error2' ``` this is due to the fact that there are 2 returns in case of a fenced code block, one with the ``endcode` but with the detection of the fenced code block (`isFencedCodeBlock`) it is not "removed".
* | | Merge pull request #8052 from albert-github/feature/issue_8051Dimitri van Heesch2020-09-241-2/+2
|\ \ \ | | | | | | | | #8051 Consider documenting debug options of doxygen
| * | | #8051 Consider documenting debug options of doxygenalbert-github2020-09-231-2/+2
| | |/ | |/| | | | | | | | | | - don't create an error in case of no argument, just handle it as a request for help. - give the usage also in case of an error (besides the error message)
* | | issue #8053 error: Attribute target redefined in SVGalbert-github2020-09-241-0/+1
|/ / | | | | | | regression on #7706 (although it has nothing to do with double runs in my opinion as indicated in the title of that issue: " Md5 hash does not match for two different runs") looks like this statement should be present.
* | Fix compiler warning when printing number of cache hits & misses (part 2)Dimitri van Heesch2020-09-211-1/+2
| |
* | issue #8037: Links using @ref stopped working in doxygen 1.8.19Dimitri van Heesch2020-09-219-2/+10
| |
* | Fix compiler warning when printing number of cache hits & missesDimitri van Heesch2020-09-211-1/+1
| |
* | Refactoring: replaced QCache by STL based LRU cache implementation.Dimitri van Heesch2020-09-205-2004/+2166
|/ | | | | Also prepared some code for multi-threaded use, and Removed tabs and trailing spaces for code.l
* Fortran comments with C comments inside (#8000)albert-github2020-09-181-1/+1
| | | | | | | | | | | | | When having a (stripped down) example like: ``` ! !!/*T !T*/ ``` we get a warning like: ``` .../ex11f.F90:4: warning: Reached end of file while still inside a (nested) comment. Nesting level 1 (probable line reference: 2, 2) ``` even though Fortran has no nested comments, and certainly doesn't have `/*` as comment signs. So message can be ignored.
* Line count mismatch for Python (#8041)albert-github2020-09-182-4/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When having an example like: ``` ## General \PYgen0 docu # # # # # General detail \PYgen1 text # and more STYLE_TEMPLATE = r''' \makeatletter ''' ## General \PYgen2 docu # # # # # # General detail \PYgen3 text # and more STYLE_TEMPLATE1 = r''' \makeatletter ''' ``` with ``` EXTRACT_ALL = YES QUIET = YES ``` we get the warnings (removed doubles and sorted): ``` bb.py:1: warning: Found unknown command '\PYgen0' bb.py:10: warning: Found unknown command '\PYgen1' bb.py:12: warning: Found unknown command '\PYgen2' bb.py:23: warning: Found unknown command '\PYgen3' ``` we see here (especially with `PYgen3` a wrong line number. After fixing the line count in `pyscanner.l` there was for the second block an offset of 1. This was caused by `commentscan.l` introduced with #7960 after correcting this this offset was also gone. Revisiting the original example of #7960 showed that here here was also an offset of 1 (probably due to a misinterpretation of where the `<tr>` warning should be mentioned), here it is now also correct.. (The original mismatch was shown of Fossies for the Pygments package)
* Fix too aggressive hyphenation of abbr. words. (#8026)Dmitriy Dorofeev2020-09-171-1/+1
| | | | | * Fix too aggressive giphenation of abbr. words. * Remove unwanted change at line 2266
* Merge pull request #8033 from albert-github/feature/bug_ftn_recognDimitri van Heesch2020-09-177-94/+76
|\ | | | | Incorrect duplicate code for Fortran fixed/free recognition
| * Incorrect duplicate code for Fortran fixed/free recognitionalbert-github2020-09-157-94/+76
| | | | | | | | There were 2 routines to recognize whether Fortran code was Fixed of Free format code, though the version in `commentcnv.l` didn't take the settings of `EXTENSION_MAPPING` into account which might lead to incorrect recognition of the format, this has been corrected.
* | Merge branch 'master' of github.com:doxygen/doxygenDimitri van Heesch2020-09-172-5/+5
|\ \
| * | Proposed fix for issue #7898 (#8029)luca-aep2020-09-162-5/+5
| | | | | | | | | | | | | | | * Proposed fix for issue #7898 - Internal search engine produce ".html" pages instead of using HTML_FILE_EXTENSION * Applied albert-github suggestions to proposed fix for issue #7898
* | | issue #8034: doxygen crashesDimitri van Heesch2020-09-172-4/+4
|/ /
* | Renamed EXTRACT_ANON_ARGUMENTS to RESOLVE_UNNAMED_PARAMS and enabled it by ↵Dimitri van Heesch2020-09-163-15/+19
| | | | | | | | default
* | Merge branch 'arg_name_as_comment' of https://github.com/tolnaisz/doxygen ↵Dimitri van Heesch2020-09-163-1/+72
|\ \ | | | | | | | | | into tolnaisz-arg_name_as_comment
| * | Support commented out argument names in declarations controlled by new cofig ↵Szabi Tolnai2020-08-113-1/+72
| | | | | | | | | | | | EXTRACT_ANON_ARGUMENTS.
* | | Merge pull request #8004 from albert-github/feature/bug_listendDimitri van Heesch2020-09-161-0/+2
|\ \ \ | | | | | | | | Warning about end of list in brief description after alias `^^` replacement
| * | | Warning about end of list in brief description after alias `^^` replacementalbert-github2020-09-031-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | With the alias: ``` ALIASES += "sbl_add_package_main_class{2}=\brief \1 ^^ \2" ``` and the comment: ``` \sbl_add_package_main_class{Defines, \details dihedrals} ``` we will get the replacement: ``` \brief Defines \ilinebr \details dihedrals ``` but this leads to a number of warnings like: ``` warning: End of list marker found without any preceding list items ``` As the end of the brief description is here `\ilinebr` this is replaced in the definition by `\ilinebr.` (so with a `.`). We first have to strip the aritficial newlines at (the beginning and) the end end of the brief description before adding the `.`. (Found as side effect of https://stackexchange.com/filters/57710/doxygen).
* | | | Merge pull request #8003 from albert-github/feature/bug_aliasDimitri van Heesch2020-09-161-6/+6
|\ \ \ \ | | | | | | | | | | Command \ilinebr remained
| * | | | Command \ilinebr reamainedalbert-github2020-09-031-6/+6
| |/ / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When having the alias: ``` ALIASES += sbl_add_package_main_class{4}="\addtogroup \1-ref-manual ^^ @{ ^^ \class \2 ^^ \brief \3 \4 ^^ @}" ``` for the comment block: ``` /** \sbl_add_package_main_class{Molecular_potential_energy, T_Phi_psi_dihedral_angles_visitor, Defines iterator over all successive pairs of dihedrals., \details Defines iterator over all successive pairs of dihedrals. \tparam CovalentStructure The type of the input covalent structure } */ ``` we get the warning: ``` warning: the name '\ilinebr' supplied as the argument of the \class, \struct, \union, or \include command is not an input file ``` The problem was that the rule: ``` <ClassDocArg2>{FILE}|"<>" { // second argument; include file ``` took the `\ilinebr`, the rule: ``` <ClassDocArg2>{DOCNL} { ``` should habe been checked before. (reported as: https://stackoverflow.com/questions/63710433/doxygen-alias-with-arguments-the-classical-linebr-problem)
* | | | Allow building doxygen as a subprojectAnonymous Maarten2020-09-161-17/+17
| |_|/ |/| |
* | | Refactoring: making xmlcode.l reentrantDimitri van Heesch2020-09-143-295/+331
| | |
* | | Refactoring: making sqlcode.l reentrantDimitri van Heesch2020-09-132-157/+150
| | |
* | | Refactoring: making vhdlcode.l reentrantDimitri van Heesch2020-09-135-1156/+1185
| | | | | | | | | | | | | | | Also fixed a few VHDL related memory leaks and one newly introduced Fortran memory leak.
* | | Merge pull request #8014 from albert-github/feature/issue_6442Dimitri van Heesch2020-09-121-0/+22
|\ \ \ | | | | | | | | issue #6442 C++: Trailing return type syntax + void
| * | | issue #6442 C++: Trailing return type syntax + voidalbert-github2020-09-081-2/+2
| | | | | | | | | | | | | | | | corrected typo
| * | | issue #6442 C++: Trailing return type syntax + voidalbert-github2020-09-071-0/+22
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When having: ``` /*! * \brief Performs some side effect */ auto side_effect_after() -> void {} ``` We get the warning: ``` warning: return type of member side_effect_after is not documented ``` as the trailing return type was not taken into consideration.
* | | | Merge branch 'albert-github-feature/issue_8015'Dimitri van Heesch2020-09-111-350/+364
|\ \ \ \
| * | | | Split lexer rulesDimitri van Heesch2020-09-111-353/+361
| | | | | | | | | | | | | | | | | | | | | | | | | Also replaced tabs by spaces and removed trailing whitespace And fixed line counting issue.
| * | | | Merge branch 'feature/issue_8015' of ↵Dimitri van Heesch2020-09-111-0/+6
| |\ \ \ \ | | | | | | | | | | | | | | | | | | https://github.com/albert-github/doxygen into albert-github-feature/issue_8015
| | * | | | issue #8015 Special command \skip and \until no longer functional in ALIASESalbert-github2020-09-071-0/+6
| | |/ / / | | | | | | | | | | | | | | | The pattern `\ilinebr` is actually also a line break, but was not handled.
* | | | | Forgot to remove some dead codeDimitri van Heesch2020-09-111-14/+0
|/ / / /
generated by cgit v0.12 at 2024-11-18 01:21:31 (GMT)