Improving insta post-infos & user-infos

Fix #962
medialab · Apr 16, 2024 · 67c5372 · 67c5372
1 parent 45171ed
commit 67c5372
Show file tree

Hide file tree

Showing 4 changed files with 20 additions and 19 deletions.
diff --git a/docs/cli.md b/docs/cli.md
@@ -4156,7 +4156,7 @@ Usage: minet instagram post-infos [-h] [-c COOKIE] [--rcfile RCFILE] [--silent]
                                   [--refresh-per-second REFRESH_PER_SECOND]
                                   [--simple-progress] [-i INPUT]
                                   [--explode EXPLODE] [-s SELECT]
-                                  [--total TOTAL] [-o OUTPUT]
+                                  [--total TOTAL] [--resume] [-o OUTPUT]
                                   post_or_post_column
 
 # Instagram post-infos
@@ -4206,6 +4206,8 @@ Optional Arguments:
   -o, --output OUTPUT           Path to the output file. Will consider `-` as
                                 stdout. If not given, results will also be
                                 printed to stdout.
+  --resume                      "Whether to resume from an aborted collection.
+                                Need -o to be set.
   --rcfile RCFILE               Custom path to a minet configuration file. More
                                 info about this here:
                                 https://github.com/medialab/minet/blob/master/do
@@ -4490,7 +4492,7 @@ Usage: minet instagram user-infos [-h] [-c COOKIE] [--rcfile RCFILE] [--silent]
                                   [--refresh-per-second REFRESH_PER_SECOND]
                                   [--simple-progress] [-i INPUT]
                                   [--explode EXPLODE] [-s SELECT]
-                                  [--total TOTAL] [-o OUTPUT]
+                                  [--total TOTAL] [--resume] [-o OUTPUT]
                                   user_or_user_column
 
 # Instagram user-infos
@@ -4542,6 +4544,8 @@ Optional Arguments:
   -o, --output OUTPUT           Path to the output file. Will consider `-` as
                                 stdout. If not given, results will also be
                                 printed to stdout.
+  --resume                      "Whether to resume from an aborted collection.
+                                Need -o to be set.
   --rcfile RCFILE               Custom path to a minet configuration file. More
                                 info about this here:
                                 https://github.com/medialab/minet/blob/master/do

diff --git a/minet/cli/instagram/__init__.py b/minet/cli/instagram/__init__.py
@@ -4,6 +4,8 @@
 #
 # Logic of the `insta` action.
 #
+from casanova import RowCountResumer
+
 from minet.cli.argparse import command, ConfigAction
 
 INSTAGRAM_COMMENTS_SUBCOMMAND = command(
@@ -152,6 +154,7 @@
         "item_label_plural": "post urls, post shortcodes or post ids",
     },
     select=True,
+    resumer=RowCountResumer,
 )
 
 INSTAGRAM_USER_FOLLOWING_SUBCOMMAND = command(
@@ -231,6 +234,7 @@
         "item_label": "username, user url or user id",
         "item_label_plural": "usernames, user urls or user ids",
     },
+    resumer=RowCountResumer
 )
 
 INSTAGRAM_USER_POSTS_SUBCOMMAND = command(

diff --git a/minet/cli/instagram/post_infos.py b/minet/cli/instagram/post_infos.py
@@ -6,6 +6,7 @@
 #
 from minet.cli.utils import with_enricher_and_loading_bar
 from minet.cli.instagram.utils import with_instagram_fatal_errors
+from minet.cli.loading_bar import LoadingBar
 from minet.instagram import InstagramAPIScraper
 from minet.instagram.types import InstagramPost
 from minet.instagram.exceptions import InstagramInvalidTargetError
@@ -15,20 +16,15 @@
 @with_enricher_and_loading_bar(
     headers=InstagramPost, title="Scraping infos", unit="posts"
 )
-def action(cli_args, enricher, loading_bar):
+def action(cli_args, enricher, loading_bar: LoadingBar):
     client = InstagramAPIScraper(cookie=cli_args.cookie)
 
-    for i, row, user in enricher.enumerate_cells(
-        cli_args.column, with_rows=True, start=1
-    ):
+    for row, user in enricher.cells(cli_args.column, with_rows=True):
         with loading_bar.step():
             try:
                 result = client.post_infos(user)
-
                 enricher.writerow(row, result)
 
             except InstagramInvalidTargetError:
-                loading_bar.print(
-                    "Given post (line %i) is probably not an Instagram post: %s"
-                    % (i, user)
-                )
+                loading_bar.inc_stat("not-found", style="warning")
+                enricher.writerow(row)
diff --git a/minet/cli/instagram/user_infos.py b/minet/cli/instagram/user_infos.py
@@ -5,6 +5,7 @@
 # Logic of the `instagram user-infos` action.
 #
 from minet.cli.utils import with_enricher_and_loading_bar
+from minet.cli.loading_bar import LoadingBar
 from minet.cli.instagram.utils import with_instagram_fatal_errors
 from minet.instagram import InstagramAPIScraper
 from minet.instagram.types import InstagramUserInfo
@@ -15,19 +16,15 @@
 @with_enricher_and_loading_bar(
     headers=InstagramUserInfo, title="Scraping infos", unit="users"
 )
-def action(cli_args, enricher, loading_bar):
+def action(cli_args, enricher, loading_bar: LoadingBar):
     client = InstagramAPIScraper(cookie=cli_args.cookie)
 
-    for i, row, user in enricher.enumerate_cells(
-        cli_args.column, with_rows=True, start=1
-    ):
+    for row, user in enricher.cells(cli_args.column, with_rows=True):
         with loading_bar.step():
             try:
                 info = client.user_infos(user)
                 enricher.writerow(row, info)
 
             except InstagramInvalidTargetError:
-                loading_bar.print(
-                    "Given user (line %i) is probably not an Instagram user: %s"
-                    % (i, user)
-                )
+                loading_bar.inc_stat("not-found", style="warning")
+                enricher.writerow(row)