diff options
| author | Terry Truong <terry06890@gmail.com> | 2023-02-05 13:55:00 +1100 |
|---|---|---|
| committer | Terry Truong <terry06890@gmail.com> | 2023-02-05 13:55:00 +1100 |
| commit | ff638e269d439c853b5182b68ff10777f12597f1 (patch) | |
| tree | 2cc62068a501884821969b6ebeba3686df8f4044 /backend/hist_data/enwiki/gen_dump_index_db.py | |
| parent | cc79c17fbc05bddc8b08f2734e721bc241123a4e (diff) | |
Use relative imports between data generation scriptspackage-imports
Avoids the need for code that modifies sys.path, but requires
running of the scripts using 'python -m' with backend/ as cwd.
Also expects constants like DB_FILE to be non-relative, due to
running from backend/.
Diffstat (limited to 'backend/hist_data/enwiki/gen_dump_index_db.py')
| -rwxr-xr-x | backend/hist_data/enwiki/gen_dump_index_db.py | 6 |
1 files changed, 4 insertions, 2 deletions
diff --git a/backend/hist_data/enwiki/gen_dump_index_db.py b/backend/hist_data/enwiki/gen_dump_index_db.py index 8872171..e4a0fbe 100755 --- a/backend/hist_data/enwiki/gen_dump_index_db.py +++ b/backend/hist_data/enwiki/gen_dump_index_db.py @@ -11,8 +11,10 @@ import re import bz2 import sqlite3 -INDEX_FILE = 'enwiki-20220501-pages-articles-multistream-index.txt.bz2' # Had about 22e6 lines -DB_FILE = 'dump_index.db' +ENWIKI_DIR = os.path.dirname(os.path.realpath(__file__)) +INDEX_FILE = os.path.join(ENWIKI_DIR, 'enwiki-20220501-pages-articles-multistream-index.txt.bz2') + # Had about 22e6 lines +DB_FILE = os.path.join(ENWIKI_DIR, 'dump_index.db') def genData(indexFile: str, dbFile: str) -> None: if os.path.exists(dbFile): |