Rename Query -> VideoQuery

CybersecurityForDemocracy · Jul 4, 2024 · 64b0044 · 64b0044
1 parent 9d5d97a
commit 64b0044
Show file tree

Hide file tree

Showing 7 changed files with 40 additions and 41 deletions.
diff --git a/README.md b/README.md
@@ -27,9 +27,9 @@ A query is a combination of a "type (and, or, not)" with multiple Conditions ("C
 
 Each condition is a combination of a "field" (Fields, F), "value" and a operation ("Operations", "Op").
 ```python
-from tiktok_research_api_helper.query import Query, Cond, Fields, Op
+from tiktok_research_api_helper.query import VideoQuery, Cond, Fields, Op
 
-query = Query(
+query = VideoQuery(
         and_=[
             Cond(Fields.hashtag_name, "garfield", Op.EQ),
             Cond(Fields.region_code, "US", Op.EQ),
@@ -45,7 +45,7 @@ query = Query(
 ```python
 from pathlib import Path
 from datetime import datetime
-from tiktok_research_api_helper.query import Query, Cond, Fields, Op
+from tiktok_research_api_helper.query import VideoQuery, Cond, Fields, Op
 from tiktok_research_api_helper.api_client import ApiClientConfig, TikTokApiClient
 
 config = ApiClientConfig(query=query,
@@ -77,9 +77,9 @@ from tiktok_research_api_helper.api_client import TikTokApiRequestClient, TikTok
 
 # reads from secrets.yaml in the same directory
 request_client = TikTokApiRequestClient.from_credentials_file(Path("./secrets.yaml"))
-from tiktok_research_api_helper.query import Query, Cond, Fields, Op
+from tiktok_research_api_helper.query import VideoQuery, Cond, Fields, Op
 
-query = Query(or_=Cond(Fields.video_id, ["7345557461438385450", "123456"], Op.IN))
+query = VideoQuery(or_=Cond(Fields.video_id, ["7345557461438385450", "123456"], Op.IN))
 
 # sample query
 req = TikTokVideoRequest(

diff --git a/src/tiktok_research_api_helper/api_client.py b/src/tiktok_research_api_helper/api_client.py
@@ -19,7 +19,7 @@
 
 from tiktok_research_api_helper import utils
 from tiktok_research_api_helper.models import Crawl, upsert_videos
-from tiktok_research_api_helper.query import Query, QueryJSONEncoder
+from tiktok_research_api_helper.query import VideoQuery, VideoQueryJSONEncoder
 
 ALL_VIDEO_DATA_URL = "https://open.tiktokapis.com/v2/research/video/query/?fields=id,video_description,create_time,region_code,share_count,view_count,like_count,comment_count,music_id,hashtag_names,username,effect_ids,voice_to_text,playlist_id"
 ALL_USER_INFO_DATA_URL = "https://open.tiktokapis.com/v2/research/user/info/?fields=display_name,bio_description,avatar_url,is_verified,follower_count,following_count,likes_count,video_count"
@@ -96,7 +96,7 @@ class TikTokApiClientFetchResult:
 # and/or comments for videos from latest crawl
 @attrs.define
 class ApiClientConfig:
-    video_query: Query
+    video_query: VideoQuery
     start_date: datetime
     end_date: datetime
     engine: Engine
@@ -119,7 +119,7 @@ class TikTokVideoRequest:
     The start date is inclusive but the end date is NOT.
     """
 
-    query: Query
+    query: VideoQuery
     start_date: str
     end_date: str  # The end date is NOT inclusive!
     max_count: int = 100
@@ -151,7 +151,7 @@ def as_json(self, indent=None):
 
         if self.cursor is not None:
             request_obj["cursor"] = self.cursor
-        return json.dumps(request_obj, cls=QueryJSONEncoder, indent=indent)
+        return json.dumps(request_obj, cls=VideoQueryJSONEncoder, indent=indent)
 
 
 @attrs.define
@@ -689,7 +689,7 @@ def api_results_iter(self) -> TikTokApiClientFetchResult:
             if not api_response.videos and crawl.has_more:
                 logging.log(
                     logging.ERROR,
-                    "No videos in response but there's still data to Crawl - Query: "
+                    "No videos in response but there's still data to Crawl - VideoQuery: "
                     f"{self._config.video_query} \n api_response.data: {api_response.data}",
                 )
             if self._config.stop_after_one_request:

diff --git a/src/tiktok_research_api_helper/cli/main.py b/src/tiktok_research_api_helper/cli/main.py
@@ -49,8 +49,8 @@
     Cond,
     Fields,
     Op,
-    Query,
-    QueryJSONEncoder,
+    VideoQuery,
+    VideoQueryJSONEncoder,
     generate_query,
 )
 
@@ -124,14 +124,14 @@ def test(
     utils.setup_logging_info_level()
     logging.log(logging.INFO, f"Arguments: {locals()}")
 
-    test_query = Query(
+    test_query = VideoQuery(
         and_=[
             Cond(Fields.hashtag_name, "snoopy", Op.EQ),
             Cond(Fields.region_code, "US", Op.EQ),
         ]
     )
 
-    logging.log(logging.INFO, f"Query: {test_query}")
+    logging.log(logging.INFO, f"VideoQuery: {test_query}")
 
     start_date_datetime = utils.str_tiktok_date_format_to_datetime("20220101")
     end_date_datetime = utils.str_tiktok_date_format_to_datetime("20220101")
@@ -270,7 +270,7 @@ def print_query(
         exclude_from_usernames=exclude_from_usernames,
     )
 
-    print(json.dumps(query, cls=QueryJSONEncoder, indent=2))
+    print(json.dumps(query, cls=VideoQueryJSONEncoder, indent=2))
 
 
 def make_crawl_date_window(crawl_span: int, crawl_lag: int) -> CrawlDateWindow:
@@ -503,7 +503,7 @@ def run(
             exclude_from_usernames=exclude_from_usernames,
         )
 
-    logging.log(logging.INFO, f"Query: {query}")
+    logging.log(logging.INFO, f"VideoQuery: {query}")
 
     if db_url:
         engine = get_engine_and_create_tables(db_url)

diff --git a/src/tiktok_research_api_helper/models.py b/src/tiktok_research_api_helper/models.py
@@ -32,7 +32,7 @@
     synonym,
 )
 
-from tiktok_research_api_helper.query import Query, QueryJSONEncoder
+from tiktok_research_api_helper.query import VideoQuery, VideoQueryJSONEncoder
 
 # See https://amercader.net/blog/beware-of-json-fields-in-sqlalchemy/
 MUTABLE_JSON = MutableDict.as_mutable(JSON)  # type: ignore
@@ -409,28 +409,28 @@ def __repr__(self) -> str:
 
     @classmethod
     def from_request(
-        cls, res_data: Mapping, query: Query, crawl_tags: Sequence[str] | None = None
+        cls, res_data: Mapping, query: VideoQuery, crawl_tags: Sequence[str] | None = None
     ) -> "Crawl":
         return cls(
             cursor=res_data["cursor"],
             has_more=res_data["has_more"],
             search_id=res_data["search_id"],
-            query=json.dumps(query, cls=QueryJSONEncoder),
+            query=json.dumps(query, cls=VideoQueryJSONEncoder),
             crawl_tags=({CrawlTag(name=name) for name in crawl_tags} if crawl_tags else set()),
         )
 
     # TODO(macpd): rename this to explain it's intent of being used before fetch starts
     @classmethod
     def from_query(
         cls,
-        query: Query,
+        query: VideoQuery,
         crawl_tags: Sequence[str] | None = None,
         has_more: bool = True,
         search_id: [int | None] = None,
     ) -> "Crawl":
         return cls(
             has_more=has_more,
-            query=json.dumps(query, cls=QueryJSONEncoder),
+            query=json.dumps(query, cls=VideoQueryJSONEncoder),
             search_id=search_id,
             crawl_tags=({CrawlTag(name=name) for name in crawl_tags} if crawl_tags else set()),
         )

diff --git a/src/tiktok_research_api_helper/query.py b/src/tiktok_research_api_helper/query.py
@@ -127,9 +127,9 @@ def convert_optional_cond_or_condseq_to_condseq(
     return optional_cond_or_seq
 
 
-# TODO(macpd): rename this to VideoQuery
+# TODO(macpd): rename this to VideoVideoQuery
 @attrs.define
-class Query:
+class VideoQuery:
     and_: OptionalCondOrCondSeq = attrs.field(
         default=None, converter=convert_optional_cond_or_condseq_to_condseq
     )
@@ -155,10 +155,9 @@ def as_dict(self):
         return formatted_operands
 
 
-# TODO(macpd): rename this to VideoQueryJSONEncoder
-class QueryJSONEncoder(json.JSONEncoder):
+class VideoQueryJSONEncoder(json.JSONEncoder):
     def default(self, o):
-        if isinstance(o, Query):
+        if isinstance(o, VideoQuery):
             return o.as_dict()
         return super().default(o)
 
@@ -218,7 +217,7 @@ def generate_query(
     exclude_all_keywords: str | None = None,
     only_from_usernames: str | None = None,
     exclude_from_usernames: str | None = None,
-) -> Query:
+) -> VideoQuery:
     query_args = {_QUERY_AND_ARG_NAME: [], _QUERY_NOT_ARG_NAME: []}
 
     if include_any_hashtags:
@@ -252,7 +251,7 @@ def generate_query(
             Cond(Fields.region_code, sorted(region_codes), Op.IN)
         )
 
-    return Query(**query_args)
+    return VideoQuery(**query_args)
 
 
 # TODO(macpd): make a class for this, comment queries, etc

diff --git a/tests/test_api_client.py b/tests/test_api_client.py
@@ -390,7 +390,7 @@ def assert_has_expected_crawl_and_videos_in_database(
         crawl = crawls[0]
         assert crawl.id == fetch_result.crawl.id
         assert crawl.cursor == len(tiktok_responses) * acquisition_config.max_count
-        assert crawl.query == json.dumps(acquisition_config.video_query, cls=query.QueryJSONEncoder)
+        assert crawl.query == json.dumps(acquisition_config.video_query, cls=query.VideoQueryJSONEncoder)
         videos = all_videos(session)
         assert len(videos) == len(tiktok_responses) * len(tiktok_responses[0].videos)
         assert len(videos) == len(fetch_result.videos)

diff --git a/tests/test_query.py b/tests/test_query.py
@@ -6,8 +6,8 @@
     Cond,
     Fields,
     Op,
-    Query,
-    QueryJSONEncoder,
+    VideoQuery,
+    VideoQueryJSONEncoder,
     generate_query,
     get_normalized_hashtag_set,
     get_normalized_keyword_set,
@@ -19,7 +19,7 @@
 def mock_query_us():
     hashtags = ["hashtag", "lol", "yay"]
 
-    return Query(
+    return VideoQuery(
         and_=[
             Cond(Fields.hashtag_name, hashtags, Op.IN),
             Cond(Fields.region_code, "US", Op.EQ),
@@ -31,7 +31,7 @@ def mock_query_us():
 def mock_query_us_ca():
     hashtags = ["hashtag", "lol", "yay"]
 
-    return Query(
+    return VideoQuery(
         and_=[
             Cond(Fields.hashtag_name, hashtags, Op.IN),
             Cond(Fields.region_code, ["US", "CA"], Op.IN),
@@ -47,7 +47,7 @@ def mock_query_exclude_some_hashtags():
         "gross",
     ]
 
-    return Query(
+    return VideoQuery(
         and_=[
             Cond(Fields.hashtag_name, include_hashtags, Op.IN),
             Cond(Fields.region_code, ["US", "CA"], Op.IN),
@@ -60,7 +60,7 @@ def mock_query_exclude_some_hashtags():
 
 @pytest.fixture
 def mock_query_create_date():
-    return Query(
+    return VideoQuery(
         and_=[
             Cond(Fields.create_date, "20230101", Op.EQ),
         ],
@@ -81,14 +81,14 @@ def test_query_create_date(mock_query_create_date):
 
 def test_query_invalid_create_date():
     with pytest.raises(ValueError):
-        Query(
+        VideoQuery(
             and_=[
                 Cond(Fields.create_date, "2023-01-01", Op.EQ),
             ],
         )
 
     with pytest.raises(ValueError):
-        Query(
+        VideoQuery(
             and_=[
                 Cond(Fields.create_date, "It's not a date", Op.EQ),
             ],
@@ -172,15 +172,15 @@ def test_query_exclude_some_hashtags(mock_query_exclude_some_hashtags):
 
 def test_invalid_region_code():
     with pytest.raises(ValueError):
-        Query(
+        VideoQuery(
             and_=[
                 Cond(Fields.region_code, "invalid", Op.EQ),
             ],
         )
 
 
 def test_query_json_decoder_us(mock_query_us):
-    assert json.dumps(mock_query_us, cls=QueryJSONEncoder, indent=1) == (
+    assert json.dumps(mock_query_us, cls=VideoQueryJSONEncoder, indent=1) == (
         """
 {
  "and": [
@@ -207,7 +207,7 @@ def test_query_json_decoder_us(mock_query_us):
 
 
 def test_query_json_decoder_us_ca(mock_query_us_ca):
-    assert json.dumps(mock_query_us_ca, cls=QueryJSONEncoder, indent=1) == (
+    assert json.dumps(mock_query_us_ca, cls=VideoQueryJSONEncoder, indent=1) == (
         """
 {
  "and": [
@@ -235,7 +235,7 @@ def test_query_json_decoder_us_ca(mock_query_us_ca):
 
 
 def test_query_json_decoder_exclude_some_hashtags(mock_query_exclude_some_hashtags):
-    assert json.dumps(mock_query_exclude_some_hashtags, cls=QueryJSONEncoder, indent=1) == (
+    assert json.dumps(mock_query_exclude_some_hashtags, cls=VideoQueryJSONEncoder, indent=1) == (
         """
 {
  "and": [