Very interesting was the exception swallowing in PackagePropertiesPart. When it was properly thrown, it already led to various errors in the junits test - I've fixed the handling for at least the ones which are in our test set
git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@1706169 13f79535-47bb-0310-9956-ffa450edef68
https://svn.apache.org/repos/asf/poi/trunk
........
r618680 | nick | 2008-02-05 16:51:29 +0100 (Tue, 05 Feb 2008) | 1 line
Add java.lang.Iterable style methods for iterating over rows and cells, but don't actually implement that, as it wasn't in jdk 1.3 or jdk 1.4, with is rather tedious
........
r618690 | nick | 2008-02-05 17:21:21 +0100 (Tue, 05 Feb 2008) | 1 line
Tweaks to the iterator use guides for hssf
........
r618692 | nick | 2008-02-05 17:32:45 +0100 (Tue, 05 Feb 2008) | 1 line
Few tweaks to the patching guide
........
r618940 | yegor | 2008-02-06 10:47:13 +0100 (Wed, 06 Feb 2008) | 1 line
3.0.2-FINAL released
........
r619001 | nick | 2008-02-06 15:35:05 +0100 (Wed, 06 Feb 2008) | 1 line
Implement hashCode and equals for HSSFFont and HSSFCellStyle
........
r619310 | yegor | 2008-02-07 09:56:59 +0100 (Thu, 07 Feb 2008) | 1 line
support for excel hypelrinks
........
r619382 | nick | 2008-02-07 13:39:12 +0100 (Thu, 07 Feb 2008) | 1 line
Patch from bug #44373 - Have HSSFDateUtil.isADateFormat support more date formats
........
r619502 | nick | 2008-02-07 17:53:23 +0100 (Thu, 07 Feb 2008) | 1 line
Handle timezones better with cell.setCellValue(Calendar), so now 20:00-03:00, 20:00+00:00 and 20:00+03:00 will all be recorded as 20:00, and not 17:00 / 20:00 / 23:00 (pass a Date not a Calendar for old behaviour) - patch from bug #38641
........
r619509 | nick | 2008-02-07 18:12:09 +0100 (Thu, 07 Feb 2008) | 1 line
Add (disabled) failing testcase for bug #44375
........
r619848 | klute | 2008-02-08 12:55:43 +0100 (Fri, 08 Feb 2008) | 1 line
- Fixed bug 44375 - HPSF now copes with a broken dictionary in Document Summary Information stream. RuntimeExceptions that occured when trying to read bogus data are now caught. Dictionary entries up to but not including the bogus one are preserved, the rest is ignored.
........
r619849 | klute | 2008-02-08 12:56:11 +0100 (Fri, 08 Feb 2008) | 1 line
- Fixed bug 44375 - HPSF now copes with a broken dictionary in Document Summary Information stream. RuntimeExceptions that occured when trying to read bogus data are now caught. Dictionary entries up to but not including the bogus one are preserved, the rest is ignored.
........
r619851 | nick | 2008-02-08 13:02:10 +0100 (Fri, 08 Feb 2008) | 1 line
Copy Rainer's change note from changes.xml to status.xml, as it's needed in both
........
r619968 | nick | 2008-02-08 19:17:29 +0100 (Fri, 08 Feb 2008) | 1 line
Tweak layout
........
r620341 | nick | 2008-02-10 22:54:13 +0100 (Sun, 10 Feb 2008) | 1 line
Test for bug #42564 - appears to already be fixed
........
r620557 | nick | 2008-02-11 18:55:09 +0100 (Mon, 11 Feb 2008) | 1 line
Another test relating to bug #42564 - this one is still failing
........
r620582 | nick | 2008-02-11 20:14:04 +0100 (Mon, 11 Feb 2008) | 1 line
If we have a document with a hpsf stream that exists, but is of the wrong type, then log a warning but continue
........
git-svn-id: https://svn.apache.org/repos/asf/poi/branches/ooxml@627544 13f79535-47bb-0310-9956-ffa450edef68
- Fixed bug 44375 - HPSF now copes with a broken dictionary in Document Summary Information stream. RuntimeExceptions that occured when trying to read bogus data are now caught. Dictionary entries up to but not including the bogus one are preserved, the rest is ignored.
Consistency check added when reading property set stream. An IllegalPropertyDataException is thrown if the property set stream is larger than it claims to be.
Applying the copy2license.pl script (with tiny modification to allow for more whitespace than it expects in the POI header) to all files. ant jar succeeds, and the svn diff has been verified by eye. Still more files to do, this is the first pass.
* Writing support added to the SummaryInformation and DocumentSummaryInformation classes. These classes now have methods for setting and removing properties. Coherent extensions are:
** Documentation section about writing standard properties added to the HPSF HOW-TO.
** Example application added showing how to modify the document summary information.
** Testcases added for testing modifying summary information and document summary information.
** PropertySetFactory extended to create SummaryInformation and DocumentSummaryInformation instances.
* Added MutablePropertySet.write(DirectoryEntry, String) to ease writing a property set to a POI filesystem document.
* Improved codepage handling.
* Bug fixed: Integral values were read and written as unsigned instead of signed.
* Reworked the mapping between variant types and Java types: Variant.VT_I4 is mapped to Integer now and Variant.VT_I8 to Long. This might cause incompatibilities if you are doing low-level HPSF programming.
* Changed SectionIDMap.DOCUMENT_SUMMARY_INFORMATION_ID from a byte[] to a byte[][] in order to contain the format ID of the first and the second section. This is an incompatible change!
* Added PropertySet.getFirstSection(). This method is similar to getSingleSection() won't choke if the property set has more than one section.
* Support for low-level reading and writing of Variant.VT_I8 type properties added.
* Unnecessary casts removed.
* Poibrowser's display format changed slightly.
git-svn-id: https://svn.apache.org/repos/asf/jakarta/poi/trunk@382887 13f79535-47bb-0310-9956-ffa450edef68
Fixed a bug where the logical comparison of two sections returned false only because one of them had property 1 (codepage) and the other had not. Now two sections are equal if their property values are equal disregarding property 1.
- Fix: HPSF now ensures that string properties are always written with the type Variant.VT_LPWSTR instead of Variant.VT_LPSTR if the codepage is 1200 (Unicode).
- Code cleanup (propertyCount in MutableSection removed).
- Codepage numbers aggregated in the new class org.apache.poi.hpsf.Constants.
git-svn-id: https://svn.apache.org/repos/asf/jakarta/poi/trunk@353561 13f79535-47bb-0310-9956-ffa450edef68
These changes introduce the following HPSF enhancements and fixes:
- Section dictionaries (aka custom properties) can be written now.
- Constructor MutableProperty(PropertySet) sets the class ID correctly now.
- Possible invalid section count fixed
- More testcases
- Cosmetics
git-svn-id: https://svn.apache.org/repos/asf/jakarta/poi/trunk@353360 13f79535-47bb-0310-9956-ffa450edef68
- PropertySetFactory.create(InputStream) no longer throws an UnexpectedPropertySetTypeException.
- HPSFRuntimeException is thrown when the application tries to read a non-integer property with Section.getPropertyIntValue(long).
- Package description updated.
- ClassIDhashCode() implemented.
git-svn-id: https://svn.apache.org/repos/asf/jakarta/poi/trunk@353349 13f79535-47bb-0310-9956-ffa450edef68
- Fixed a bug that occured when reading a section with properties that are not stored with ascending offsets in the property list.
- An error message is printed to standard error now when an unsupported variant type occurs. The data bytes are still available.
git-svn-id: https://svn.apache.org/repos/asf/jakarta/poi/trunk@353307 13f79535-47bb-0310-9956-ffa450edef68
Finished my work converting HPSF to use the org.apache.poi.util.LittleEndian
implementation. ClassID and poibrowser/PropertySetDescriptorRendeder are probably broken. I couldn't figure out what ClassID is meaning to do someone else
who understands the code there will have to work on it.
PR:
Obtained from:
Submitted by:
Reviewed by:
git-svn-id: https://svn.apache.org/repos/asf/jakarta/poi/trunk@352639 13f79535-47bb-0310-9956-ffa450edef68