GetCanKnowMove should be checking for Smeargle knowing thousand arrows/waves.
GetNeededMoves should also return 1ka/1kw for needed moves, and then no encounters.
All uses of InvalidSketch are behind that method, so any future Smeargle modifications (like sketching a dummied move) will be checked (yay BDSP considerations).
Removes duplicates if every encounter in the same index is the same as a
previous index. The following were affected, and only the first index
was retained.
- 5, 7 (Index 2 has 1 move different on 1 mon)
- 3, 4
- 11, 13
- 16, 18
- 27, 28, 30, 32
- 33, 35
- 36, 38
- 42, 44
- 45, 47
- 48, 50, 52
- 53, 55, 57
- 59, 61
Just allocate the parent array and pass it to HashSet; having an ICollection, prevents resizing repeatedly and gc'ing the temp array is cheap.
Actually, only create a hashset if we're adding a gift from external DB. Saves on that allocation, and keeps the final result as an array (fastest iterating).
Removes 1.5MB of allocation; we don't need to preserve the encrypted state.
Revise savefile clone operation to use its own constructor, as we need to call the base constructor first (backup file clone...) prior to decrypting.
Expose CryptStaticXorpadBytes as GetDecryptedRawData no longer exists.
Regex objects aren't cheap, and the RegexCache has DefaultMaxCacheSize = 15.
We're checking 4,000 regexes for each unique string, so just keep the created regexes around instead of cycling new through the cache.
+4MB passive consumption, but each IsFiltered call no longer generates >4MB of discarded objects.
My unit tests run >25% faster now... nice?!
handle all permutations of Indeedee-M/F (M/F)
if it detects (F), it uses female, otherwise it uses whatever the detected form was (0/1) and sets the values accordingly.
hovering would grab ILocation, which didn't use the flexible met location fetch.
Just repoint stuff and have it as a virtual call, since most of these are fixed values, there's no point increasing the size of the classes besides those for gen4.
Add dragonite to the uncapturable list, removes catchrate comparison for dragonite since it's filtered out.
Skip some linq by doing the loop directly in the method. Replace local methods with captured variables with direct if-returns.
Looks like Mr. Rime case wasn't being handled, so I rewrote it. Better performance, less complexity. No need to double-reference the moves.
Cache a single Valid evolution result; every parse can reuse that object.
Only remaining uses of regex in the sln are wordfilter (required), and unescape for QR replies from web api.
This is slightly faster -- cover all cases if the response json changes (shouldn't).
More in line with modernizing the codebase with latest c# syntax
improve web-qr decode speed slightly (no linq skiptake)
get money/coin mask without a temporary string (lol performance)
Hardcode egg names, don't rely on string resource loading order for SpeciesNames to get an un-mutated copy.
fixes the infrequent race condition for legality unit tests
O(n) for speed & memory usage, as good as we can get.
Modified method is now less dumb. Array Copy doesn't have to worry about same-array overlapping src/dest regions
The EncDB doesn't specify form; the encounter methods verify Form.
This is actually a fix; pkm that can change forms aren't really encounter trades, but we shouldn't be strict matching here.
Rotom Bike water upgrade isn't available until you have sufficient progress, so you can't get these dens with the lowest rank.
Co-Authored-By: Lusamine <30205550+Lusamine@users.noreply.github.com>
Nice bug ya got there, gamefreak.
Coulda cleared the AffixedRibbon value instead of copying it on Shedinja creation, and it would have made this unnecessary.
Please ditch the Affixed Ribbon gimmick for future games, thanks!
Distribution raid encounters from before each of the DLC patches can only be captured in existing Wild Area locations.
Store the BCAT index in the encounter details and use it to compare when the BCAT was available.
Switch expression had an issue with interpreting the object's type (unboxing, yay)
should probably just return a typed object from the parse, but nah, this is a quick fix.
Farfetch’d uses the non-vertical apostrophe, still.
Showdown is consistent with their usage (in/outputting ’), but we'll keep the '->’ for that sanitization just in case other services/users enter the wrong char.
Closes#3185
Closes#3194
Exports are fine, imports with an ambiguous language (can be either JPN or INT) while illegal don't result in the above branches passing it.
Inputs with <=6 chars for OT&Nick will be ambiguous; PKM editor.
- Settings now stored as json next to exe
- Settings now exposes all legality checking setttings that can be changed
- Slot hovering now can play cries in MGDB/PKMDB/etc, not just the main boxes.
- Enhanced hover text for mystery gifts and encounters that have movesets
- Show recently loaded save files in ctrl-f browser
- Toggle auto-load savefile setting to be none/detect-latest/last-loaded
- Custom extensions & extra backup paths can now be configured directly in the json settings
- Settings editor now uses propertygrid & tabs.
This change revises the logic for an Egg Move's RelearnMoves.
Removes a class that is no longer used
Passes the large value-passing struct byref via in keyword
Revises handling so that early-returned results from the MoveBreed processing are marked with their move source rather than leaving them unvisited.
The moveset processor uses depth search recursion, and returns false if the move that it is processing is not possible for the hatched moveset. If the depth search never reaches the check for base moves, all moves won't be marked for their origins, resulting in the result array not being accurate to use for suggestion purposes.
Yay multi-purpose logic (validation w/early returns & fixing needing fully processed data).
