Commit 0aa9348b1fe3144e7ab0f0846d33a7d0a0c97dc3
1 parent
d6b5e4ef
reorganize
git-svn-id: svn+q:///qpdf/trunk@791 71b93d88-0707-0410-a8cf-f5a4172ac649
Showing
1 changed file
with
25 additions
and
25 deletions
TODO
| 1 | 1 | 2.1 |
| 2 | 2 | === |
| 3 | 3 | |
| 4 | - * For some reason, /MT with msvc and -static-libgcc with mingw both | |
| 5 | - produce lots of test suite failures. Why do some things work and | |
| 6 | - some things not work? Is it static data in PCRE or zlib? Is it | |
| 7 | - something in my code? | |
| 8 | - | |
| 9 | 4 | * Update documentation to reflect new command line flags and any |
| 10 | 5 | other relevant changes. Should read through ChangeLog and the |
| 11 | 6 | manual before releasing 2.1. |
| 12 | 7 | |
| 8 | + * Update release documentation to remember not to include debugging | |
| 9 | + in the Windows release. | |
| 10 | + | |
| 13 | 11 | * Add comments for the security functions that map them back to the |
| 14 | 12 | items in Adobe's products. |
| 15 | 13 | |
| ... | ... | @@ -23,41 +21,42 @@ |
| 23 | 21 | through all error messages to try to include all these fields as |
| 24 | 22 | appropriate. Make sure invalid password is specifically |
| 25 | 23 | detectable. I/O errors and so forth should also be |
| 26 | - distinguishable. | |
| 24 | + distinguishable. Make sure all errors include information about | |
| 25 | + the most recent read location including byte offset and | |
| 26 | + object/generation number. | |
| 27 | 27 | |
| 28 | - * Test DLL works on a fresh XP installation | |
| 28 | + * It might be nice to be able to trap I/O errors separately from | |
| 29 | + other errors; especially be able to separate errors that the user | |
| 30 | + can fix (like permission errors) from errors that they probably | |
| 31 | + can't fix like corrupted PDF files, unsupported filters, or | |
| 32 | + internal errors. However, only QPDF::processFile(), which does the | |
| 33 | + initial read, and QPDFWriter::QPDFWriter(), which does the initial | |
| 34 | + write, are at all likely to generate such errors for a case other | |
| 35 | + than a catastrophic failure. | |
| 29 | 36 | |
| 30 | 37 | * "Delphi wrapper unit 'qpdf.pas' created by Zarko Gajic |
| 31 | 38 | (http://delphi.about.com). .. use at your own risk and for whatever |
| 32 | 39 | the purpose you want .. no support provided. Sample code provided." |
| 33 | 40 | |
| 41 | +2.2 | |
| 42 | +=== | |
| 43 | + | |
| 34 | 44 | * Add ability to create new streams or replace stream data. Consider |
| 35 | 45 | stream data sources to include a file and offset, a buffer, or a |
| 36 | 46 | some kind of callback mechanism. Find messages exchanged with |
| 37 | 47 | Stefan Heinsen <stefan.heinsen@gmx.de> in August, 2009. He seems |
| 38 | 48 | to like to send encrypted mail. (key 01FCC336) |
| 39 | 49 | |
| 40 | - * Improve the error message generated when processing the broken | |
| 41 | - files...ideally we should provide the object number currently being | |
| 42 | - read | |
| 50 | + * See whether we can do anything with /V > 3 in the encryption | |
| 51 | + dictionary. (V = 4 is Crypt Filters.) See | |
| 52 | + ~/Q/pdf-collection/R4-encrypt-PDF_Inside_and_Out.pdf | |
| 53 | + | |
| 54 | + * Look at page splitting. | |
| 43 | 55 | |
| 44 | 56 | |
| 45 | 57 | General |
| 46 | 58 | ======= |
| 47 | 59 | |
| 48 | - * It might be nice to be able to trap I/O errors separately from | |
| 49 | - other errors; especially be able to separate errors that the user | |
| 50 | - can fix (like permission errors) from errors that they probably | |
| 51 | - can't fix like corrupted PDF files, unsupported filters, or | |
| 52 | - internal errors. However, only QPDF::processFile(), which does the | |
| 53 | - initial read, and QPDFWriter::QPDFWriter(), which does the initial | |
| 54 | - write, are at all likely to generate such errors for a case other | |
| 55 | - than a catastrophic failure. | |
| 56 | - | |
| 57 | - * See whether we can do anything with /V > 3 in the encryption | |
| 58 | - dictionary. (V = 4 is Crypt Filters.) See | |
| 59 | - ~/Q/pdf-collection/R4-encrypt-PDF_Inside_and_Out.pdf | |
| 60 | - | |
| 61 | 60 | * The second xref stream for linearized files has to be padded only |
| 62 | 61 | because we need file_size as computed in pass 1 to be accurate. If |
| 63 | 62 | we were not allowing writing to a pipe, we could seek back to the |
| ... | ... | @@ -122,5 +121,6 @@ taking care not to traverse into the pages tree, and then to fabricate |
| 122 | 121 | a new pages tree. |
| 123 | 122 | |
| 124 | 123 | Either way, care must be taken to handle other things such as |
| 125 | -outlines, page labels, thumbnails, threads, etc. in a sensible way. | |
| 126 | -This may include simply omitting information other than page content. | |
| 124 | +outlines, page labels, thumbnails, threads, zones, etc. in a sensible | |
| 125 | +way. This may include simply omitting information other than page | |
| 126 | +content. | ... | ... |