mem: exclude unused spaCy pipeline components to reduce model memory

[KRRT7]

Only tok2vec, tagger, and sentence splitting are used (pos_tag and sent_tokenize). Exclude ner, parser, lemmatizer, attribute_ruler when loading en_core_web_sm, and add lightweight sentencizer to replace the dependency parser for sentence boundary detection.

Benchmark

Measured with memray (memray run + memray stats --json), 3 rounds × 5 texts through pos_tag() + sent_tokenize() + word_tokenize(), Python 3.12.

spaCy en_core_web_sm — component exclusion benchmark
pos_tag + sent_tokenize + word_tokenize  |  3 rounds x 5 texts  |  Python 3.12.12

Configuration                               Peak MB      Saved      %
----------------------------------------------------------------------
All components (default)                    202.1MB      0.0MB   0.0%
Exclude ner/parser/lemma/attr_ruler         189.3MB     12.7MB   6.3%

[badGarnet]

The trade-off — sentence splitting quality:

• Currently sent_tokenize() (line 173) gets sentence boundaries from the parser (dependency-parse-based, more accurate).
• After this change, it uses the sentencizer (rule-based, splits on punctuation like .?!).
• This is less accurate for edge cases (abbreviations like "Dr. Smith", numbered lists, etc.) but faster and lighter.

I think this is why we see the ingest test failure (some minor changes). I would put parser back just to be safe.

[KRRT7]

Good call — updated to keep parser loaded for accurate sentence boundaries. Now only excluding ner, lemmatizer, and attribute_ruler. Memory savings drop from ~12.7 MiB to ~7 MiB but we avoid the sentence splitting regression.

Also rebased onto main and bumped to 0.22.9.