1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25node_modules/
# Downloaded PDFs and the chunks split out of them. Content-addressed and
# regenerable from the CSV, and at corpus scale this is tens of gigabytes of other
# people's documents — it does not belong in git.
cache/
# Run artifacts: one directory per document, holding status.json, log.jsonl,
# diagnostics.json and output.html. These are the raw material of the analysis and
# are worth keeping, but they are large and they contain the content of the
# documents under test. Commit a summary, not the corpus.
runs/
results.jsonl
summary.json
prepared.jsonl
corpus.jsonl
# The URL list itself. Whether a given list is publishable is a decision for
# whoever assembled it, so the default is not to.
*.csv
!example.csv
.env
.DS_Store