| - Implemented caching for latex string widths. getLatexStrWidth now calls a series of functions that check to see if the width for a particular string has been stored in a dictionary of string widths. The dictionary is created and managed using the filehash package. The code that calls LaTeX has been moved to a new function called latexParseStrForWidth and is only called if the width is not found in the filehash. |
|
More...
|
about 16 years ago
|
| - Implemented caching for latex string widths. getLatexStrWidth now calls a series of functions that check to see if the width for a particular string has been stored in a dictionary of string widths. The dictionary is created and managed using the filehash package. The code that calls LaTeX has been moved to a new function called latexParseStrForWidth and is only called if the width is not found in the filehash. |
|
More...
|
about 16 years ago
|
| - Added new default return values for TikZ_MetricInfo taken from the Quartz device. |
|
More...
|
about 16 years ago
|
| - Added new default return values for TikZ_MetricInfo taken from the Quartz device. |
|
More...
|
about 16 years ago
|
| - Added the filehash package as a dependency- it will be used to keep track of whether metrics for strings have been calculated before. - This package is now namespaced. This means that any objects accessable to the users must be explicitly exported in the NAMESPACE file. Similarly any objects from other packages that are to be used must be explicitly imported. This will allow us to shield certain options and variables from inadvertant tampering by users. |
|
More...
|
about 16 years ago
|
| - Added the filehash package as a dependency- it will be used to keep track of whether metrics for strings have been calculated before. - This package is now namespaced. This means that any objects accessable to the users must be explicitly exported in the NAMESPACE file. Similarly any objects from other packages that are to be used must be explicitly imported. This will allow us to shield certain options and variables from inadvertant tampering by users. |
|
More...
|
about 16 years ago
|
| - Oops, latexCmd is really gone now. |
|
More...
|
about 16 years ago
|
| - Oops, latexCmd is really gone now. |
|
More...
|
about 16 years ago
|
| - Removed latexCmd from the C code as there is no reason to pass it in anymore. |
|
More...
|
about 16 years ago
|
| - Removed latexCmd from the C code as there is no reason to pass it in anymore. |
|
More...
|
about 16 years ago
|
| Merge branch 'Text_Alignment' of git://github.com/Sharpie/RTikZDevice into string-width |
|
More...
|
about 16 years ago
|
| Merge branch 'Text_Alignment' of git://github.com/Sharpie/RTikZDevice into string-width |
|
More...
|
about 16 years ago
|
| - Replicated the functionality of the C code GetLateXStringWidth in a new R function called getLatexStrWidth. The C code had problems with output suppression on both Windows and Linux. This problem was also causing Windows to crash. I'm sure the issues could be sorted using a pure C approach- but it's soooooo much easier to implement this functionality in R. TikZ_StrWidth now performs a callback to R which causes LaTeX to be launched using the R tools such as system(). |
|
More...
|
about 16 years ago
|
| - Replicated the functionality of the C code GetLateXStringWidth in a new R function called getLatexStrWidth. The C code had problems with output suppression on both Windows and Linux. This problem was also causing Windows to crash. I'm sure the issues could be sorted using a pure C approach- but it's soooooo much easier to implement this functionality in R. TikZ_StrWidth now performs a callback to R which causes LaTeX to be launched using the R tools such as system(). |
|
More...
|
about 16 years ago
|
| - Removed some testing code from tikzDevice.c - Wrapped the system() call in zzz.R that tests for the existance of LaTeX in suppressWarning() in order to prevent failure spam from showing up on Windows. |
|
More...
|
about 16 years ago
|
| - Removed some testing code from tikzDevice.c - Wrapped the system() call in zzz.R that tests for the existance of LaTeX in suppressWarning() in order to prevent failure spam from showing up on Windows. |
|
More...
|
about 16 years ago
|
| Merge branch 'master' into Text_Alignment |
|
More...
|
about 16 years ago
|
| Merge branch 'master' into Text_Alignment |
|
More...
|
about 16 years ago
|
| Merge of github master branch to r-forge SVN repository. Fixed a bug involving an unallocated pointer that was causing the device to crash on Linux and Windows. |
|
More...
|
about 16 years ago
|
| Merge of github master branch to r-forge SVN repository. Fixed a bug involving an unallocated pointer that was causing the device to crash on Linux and Windows. |
|
More...
|
about 16 years ago
|
| - Fixed a bug introduced along with clipping routines in revision 80f68ff775ed90608c3f5e9fb5b9667b7dd36f49. During initialization, no memory was allocated for the pointer plotParams. This resulted in the call to StyleDef in TikZ_Clip passing a plotParams pointer that did not actually point anywhere. Strangely this only created fatal crashes on Linux and Windows... |
|
More...
|
about 16 years ago
|
| - Fixed a bug introduced along with clipping routines in revision 80f68ff775ed90608c3f5e9fb5b9667b7dd36f49. During initialization, no memory was allocated for the pointer plotParams. This resulted in the call to StyleDef in TikZ_Clip passing a plotParams pointer that did not actually point anywhere. Strangely this only created fatal crashes on Linux and Windows... |
|
More...
|
about 16 years ago
|
| - Added a robust checking code to zzz.R that determines if LaTeX can be executed. If the check fails, the package will fail to load- hopefully this will protect people from R crashing if the tikz device tries and fails to run LaTeX. Since there is no standardized variable in R that holds the path to LaTeX (of course there isn't, who would want such a thing?), several locations that claim to be "THE path to LaTeX" are checked in the following order: |
|
More...
|
about 16 years ago
|
| - Added a robust checking code to zzz.R that determines if LaTeX can be executed. If the check fails, the package will fail to load- hopefully this will protect people from R crashing if the tikz device tries and fails to run LaTeX. Since there is no standardized variable in R that holds the path to LaTeX (of course there isn't, who would want such a thing?), several locations that claim to be "THE path to LaTeX" are checked in the following order: |
|
More...
|
about 16 years ago
|
| Merge branch 'Text_Alignment' of git://github.com/Sharpie/RTikZDevice into string-width |
|
More...
|
about 16 years ago
|
| Merge branch 'Text_Alignment' of git://github.com/Sharpie/RTikZDevice into string-width |
|
More...
|
about 16 years ago
|
| - Working on legend and text alignment bugs. Most of the problem with the width of legend boxes seems to be that GetLatexStrWidth was deviding it's widths by 72. According to the documentation, widths should be reported in *device* units which are currently pixels. I think LaTeX currently reports string widths in pixels as well, dividing by 72 converts the result to inches, which are much too small and caused the narrow legend boxes. |
|
More...
|
about 16 years ago
|
| - Working on legend and text alignment bugs. Most of the problem with the width of legend boxes seems to be that GetLatexStrWidth was deviding it's widths by 72. According to the documentation, widths should be reported in *device* units which are currently pixels. I think LaTeX currently reports string widths in pixels as well, dividing by 72 converts the result to inches, which are much too small and caused the narrow legend boxes. |
|
More...
|
about 16 years ago
|
| Merge branch 'string-width' of git://github.com/cameronbracken/rtikzdevice into Text_Alignment |
|
More...
|
about 16 years ago
|
| Merge branch 'string-width' of git://github.com/cameronbracken/rtikzdevice into Text_Alignment |
|
More...
|
about 16 years ago
|
