No changes have been made to the output format or contents.
- Automatically detect game version number from executable.
- Explicitly ignore
PLYT
records to prevent crash with EcksEdit 4.1.6.9 on latest PTS versions. - Significantly speed up dump scripts (from 5 hours to 2 hours) by reducing "multiplexer" complexity and inlining utility functions.
- Modernise Python project dependency management.
- Format all Python code using Black.
- Re-format all Pascal code using updated project code style.
- Breaking change:
During recursive terminal traversal for
TERM.wiki
, terminals are no longer included more than once. The previous behavior, though useful in some cases, would result in out-of-memory errors, excessively large exports, and bad performance when exporting quiz-like terminals. - Add a centrally-kept changelog.
- Replace improper usages of
eByPath
witheByName
, and useeByPath
to combine calls. - Fix several memory leaks.
- Fix incorrect script name in
README.md
.
- Added
ALCH
dumps for ingestibles. (#54) - Fixed minor code style and performance issues.
- Breaking change: The main Python script has been renamed from
export.py
tofo76dumps.py
. - Breaking change: All paths in
config.py
(andconfig_default.py
) should now be of typePath
instead of typestr
. - Breaking change: Parameter
ba2extract_compatdata_path
inconfig.py
(andconfig_default.py
) is now interpreted as an absolute path instead of a relative path. - xEdit dump parts (such as
tabular.IDs.csv.001
) are now moved to the directorydumps/_parts/
after they have been merged together. Similarly, files that have been archived are now moved to the directorydumps/_archived/
after archiving has been completed. As a result, all files that should be attached to a release can now be selected instantly. - Warnings and errors that occur while generating and processing dumps have clearer descriptions now.
- Updated dependencies.
- New dump: Credits are dumped in
raw.credits.txt
. (#41) - New dump: Curve tables are dumped in
raw.curvetables.zip
. .ba2
files can now be processed, allowing raw files to be exported from the game.- xEdit archiving happens in the background of ba2extract.
- Dump scripts can be configured from
config.py
. - Slimmed-down README, with most information moved over to the wiki.
- ESMs are optionally archived as a utility. These archives will not be uploaded.
- Linux dump processing script now correctly creates
.ini
parent directory.
- Dump processing script now supports Linux using Proton.
- Dump post-processing has been automated using a Python script.
.7z
archives are created for dumps of at least 10MB instead of at least 10MiB.- Column names in the
.db
now contain_
where the corresponding.csv
has whitespace. - Values in the
.db
now retain all leading and trailing whitespace that are present in the game files. - All columns in the
.db
are now of typeTEXT
to prevent lossy and incorrect automatic type detection.
- Renamed all occurrences of variable
e
toel
.
- Fixed crash when a float is unexpectedly a string.
- New dumps
- Flora (
FLOR.csv
) (#20) - Leveled items (
LVLI.csv
) (#20) - Object modifications (
OMOD.csv
) (#21) - Weapons (
WEAP.csv
) (#20/#21) - Armor locations (
ARMO_LOC.csv
) (#20) - Flora locations (
FLOR_LOC.csv
) (#20) - Leveled item locations (
LVLI_LOC.csv
) (#20) - Miscellaneous item locations (
MISC_LOC.csv
) (#20) - Weapon locations (
WEAP_LOC.csv
) (#20)
- Flora (
- Dump changes
- Properties, components, and faction relations are now expressed as JSON objects instead of strings. (#27)
- References (e.g. keywords, perks, factions) are now dumped as
<edid> "<name>" [<sig>:<formid>]
("<name>"
is optional) instead of only<edid>
. (#35) - Constructible object dumps now use only one column for the product and one column for the recipe, formatted as references. (5bd2b25bfabedb8a758d4bdaf95ee06a5a870b2f)
- Bug fixes
- Empty script notes in
DIAL.csv
no longer have italics around them. (#26) - NPC perks no longer have a trailing
=
in each entry. (#28) - JSON is now escaped more properly. (#31)
- Constructible objects no longer have missing component lists. (#33)
- Empty script notes in
- New script features
- Exposed
ExportAll
script to xedit without having to run thefo76pas
. (ce17851a2beed1dbc2f639677aab3bc815ff9145) - Ability to select which dump scripts to run when using
ExportAll
. (#23) - Terminal dumps are now exported in parts, like ID exports. (#24)
- Create
_done.txt
afterExportAll
has finished. (#25) - Display error report after
ExportAll
has finished. (#29)
- Exposed
- Added image path fields for
ENTM
records. (#19)
- Added armour dump (
ARMO.csv
) (#17). - Added outfit dump (
OTFT.csv
). - Added default outfit field to
NPC_.csv
. - Added description field to
ENTM.csv
.
- Replaced
ERROR
strings in holotape exports with more extensive error messages. (#14) - Holotapes with terminal contents now have a reference to the terminal in question. (#15)
- Unreferenced terminal entries are now also exported, hence the need to compress the exported file. (#15)
- Include more information about NPCs, in particular about their appearance. (#16)
- Flattened lists have been replaced with JSON arrays in all tabular exports.
- Flattened lists are now sorted alphabetically.
- Added
FACT.csv
. - The
Item name
column inMISC.csv
has been renamed toName
. - Fixed a bug where double quotes (
"
) were not properly escaped in CSVs.
- Dumps no longer order records by base ID.
TERM.wiki
no longer features unnecessary headers.TERM.wiki
now puts the transcript template around all transcripts.- Cleaned up some scripts (#4, #5).
- Added
ExportAll.fo76pas
file to automatically run all export scripts in sequence.
- Added
ENTM.csv
,CLAS.csv
,NPC_.csv
, andRACE.csv
dumps.
- Added more information to
COBJ.csv
. - Added editor IDs to
BOOK.wiki
andNOTE.csv
.
- Added ESM name to all records in all exports.
- Export scripts now check record signature before trying to export it.
- Export scripts now follow a common style guide.
- Added
MISC.csv
. - Added
DIAL.wiki
.
- Reduced
IDs.csv
script memory consumption by outputting export in parts. - Added keywords column to
IDs.csv
. - Added
SCRAP.csv
. - Added
COBJ.csv
.
- Added
BOOK.wiki
. - Added
TERM.wiki
.
Initial release.