Lint only changed SQL files on PRs

.github/workflows/linter.yml

@@ -44,7 +44,6 @@ jobs:
           VALIDATE_MARKDOWN: true
           VALIDATE_PYTHON_PYLINT: true
           VALIDATE_PYTHON_FLAKE8: true
-          VALIDATE_SQLFLUFF: true
           VALIDATE_YAML: true
 
   dependabot:
@@ -73,7 +72,6 @@ jobs:
             steps.metadata.outputs.update-type == 'version-update:semver-minor'
           ) && (
             contains(steps.metadata.outputs.dependency-names, 'prettier') ||
-            contains(steps.metadata.outputs.dependency-names, 'sqlfluff') ||
             contains(steps.metadata.outputs.dependency-names, 'super-linter')
           )
         run: gh pr merge --admin --squash "$PR_URL"

.github/workflows/lintsql.yml

@@ -7,7 +7,8 @@ on:
   workflow_dispatch:
   pull_request:
     paths:
-      src/requirements.txt
+      - src/requirements.txt
+      - 'sql/**.sql'
 jobs:
   lint:
     name: Lint SQL
@@ -23,9 +24,46 @@ jobs:
         with:
           python-version: '3.12'
       - name: Lint SQL code
-        if: |
-            github.event_name == 'workflow_dispatch' ||
-            startsWith(github.event.pull_request.title,'Bump sqlfluff') == true
         run: |
-          pip install -r src/requirements.txt
-          sqlfluff lint sql -p 4
+          pip install -r src/requirements.txt -q
+          if [ "${{ github.event_name }}" == "workflow_dispatch" ] || \
+            [[ "${{ github.event.pull_request.title }}" == Bump\ sqlfluff* ]]; then
+            # Lint all SQL files for workflow_dispatch or sqlfluff bumps
+            sqlfluff lint sql -p 4
+          else
+            # Lint only changed SQL files
+            git diff --name-only --diff-filter=ACMRT origin/${{ github.base_ref }}...HEAD \
+              | grep '\.sql$' \
+              | xargs -r sqlfluff lint
+          fi
+
+
+  dependabot:
+    name: Dependabot auto-merge
+    runs-on: ubuntu-latest
+    needs: lint
+    if: |
+      github.event.pull_request.user.login == 'dependabot[bot]' &&
+      github.repository == 'HTTPArchive/almanac.httparchive.org'
+
+    permissions:
+      contents: write
+      pull-requests: write
+
+    steps:
+      - name: Dependabot metadata
+        id: metadata
+        uses: dependabot/fetch-metadata@v2
+        with:
+          github-token: "${{ secrets.GITHUB_TOKEN }}"
+
+      - name: Enable auto-merge for Dependabot PRs
+        if: |
+          (
+            steps.metadata.outputs.update-type == 'version-update:semver-patch' ||
+            steps.metadata.outputs.update-type == 'version-update:semver-minor'
+          ) && contains(steps.metadata.outputs.dependency-names, 'sqlfluff')
+        run: gh pr merge --admin --squash "$PR_URL"
+        env:
+          PR_URL: ${{github.event.pull_request.html_url}}
+          GH_TOKEN: ${{secrets.GITHUB_TOKEN}}

sql/2019/fonts/06_32.sql

@@ -1,23 +1,19 @@
 #standardSQL
 # 06_32: Top font hosts
 SELECT
-  *
-FROM (
-  SELECT
-    client,
-    NET.HOST(url) AS host,
-    COUNT(0) AS freq,
-    SUM(COUNT(0)) OVER (PARTITION BY client) AS total,
-    ROUND(COUNT(0) * 100 / SUM(COUNT(0)) OVER (PARTITION BY client), 2) AS pct
-  FROM
-    `httparchive.almanac.requests`
-  WHERE
-    date = '2019-07-01' AND
-    type = 'font'
-  GROUP BY
-    client,
-    host
-  ORDER BY
-    freq / total DESC
-)
+  client,
+  NET.HOST(url) AS host,
+  COUNT(0) AS freq,
+  SUM(COUNT(0)) OVER (PARTITION BY client) AS total,
+  ROUND(COUNT(0) * 100 / SUM(COUNT(0)) OVER (PARTITION BY client), 2) AS pct
+FROM
+  `httparchive.almanac.requests`
+WHERE
+  date = '2019-07-01' AND
+  type = 'font'
+GROUP BY
+  client,
+  host
+ORDER BY
+  freq / total DESC
 LIMIT 100

sql/2021/css/image_dimension_popularity.sql

@@ -1,59 +1,55 @@
 #standardSQL
 # CSS-initiated image px dimension popularity
 SELECT
-  *
+  client,
+  height,
+  width,
+  COUNT(0) AS freq,
+  SUM(COUNT(0)) OVER (PARTITION BY client) AS total,
+  COUNT(0) / SUM(COUNT(0)) OVER (PARTITION BY client) AS pct
 FROM (
   SELECT
     client,
-    height,
-    width,
-    COUNT(0) AS freq,
-    SUM(COUNT(0)) OVER (PARTITION BY client) AS total,
-    COUNT(0) / SUM(COUNT(0)) OVER (PARTITION BY client) AS pct
-  FROM (
-    SELECT
-      client,
-      page,
-      url AS img_url,
-      JSON_VALUE(payload, '$._initiator') AS css_url
-    FROM
-      `httparchive.almanac.requests`
-    WHERE
-      date = '2021-07-01' AND
-      type = 'image'
-  )
-  JOIN (
-    SELECT
-      client,
-      page,
-      url AS css_url
-    FROM
-      `httparchive.almanac.requests`
-    WHERE
-      date = '2021-07-01' AND
-      type = 'css'
-  )
-  USING (client, page, css_url)
-  JOIN (
-    SELECT
-      _TABLE_SUFFIX AS client,
-      url AS page,
-      JSON_EXTRACT_SCALAR(image, '$.url') AS img_url,
-      SAFE_CAST(JSON_EXTRACT_SCALAR(image, '$.naturalHeight') AS INT64) AS height,
-      SAFE_CAST(JSON_EXTRACT_SCALAR(image, '$.naturalWidth') AS INT64) AS width
-    FROM
-      `httparchive.pages.2021_07_01_*`,
-      UNNEST(JSON_EXTRACT_ARRAY(JSON_EXTRACT_SCALAR(payload, '$._Images'), '$')) AS image
-  )
-  USING (client, page, img_url)
+    page,
+    url AS img_url,
+    JSON_VALUE(payload, '$._initiator') AS css_url
+  FROM
+    `httparchive.almanac.requests`
   WHERE
-    height IS NOT NULL AND
-    width IS NOT NULL
-  GROUP BY
+    date = '2021-07-01' AND
+    type = 'image'
+)
+JOIN (
+  SELECT
     client,
-    height,
-    width
-  ORDER BY
-    pct DESC
+    page,
+    url AS css_url
+  FROM
+    `httparchive.almanac.requests`
+  WHERE
+    date = '2021-07-01' AND
+    type = 'css'
+)
+USING (client, page, css_url)
+JOIN (
+  SELECT
+    _TABLE_SUFFIX AS client,
+    url AS page,
+    JSON_EXTRACT_SCALAR(image, '$.url') AS img_url,
+    SAFE_CAST(JSON_EXTRACT_SCALAR(image, '$.naturalHeight') AS INT64) AS height,
+    SAFE_CAST(JSON_EXTRACT_SCALAR(image, '$.naturalWidth') AS INT64) AS width
+  FROM
+    `httparchive.pages.2021_07_01_*`,
+    UNNEST(JSON_EXTRACT_ARRAY(JSON_EXTRACT_SCALAR(payload, '$._Images'), '$')) AS image
 )
+USING (client, page, img_url)
+WHERE
+  height IS NOT NULL AND
+  width IS NOT NULL
+GROUP BY
+  client,
+  height,
+  width
+ORDER BY
+  pct DESC
 LIMIT 500

sql/2021/css/keyframes_positions.sql

@@ -24,36 +24,32 @@ try {
 ''';
 
 SELECT
-  *
-FROM (
+  client,
+  position,
+  COUNT(DISTINCT page) AS pages,
+  ANY_VALUE(total) AS total_pages,
+  COUNT(DISTINCT page) / ANY_VALUE(total) AS pct_pages,
+  COUNT(0) AS freq,
+  SUM(COUNT(0)) OVER (PARTITION BY client) AS total,
+  COUNT(0) / SUM(COUNT(0)) OVER (PARTITION BY client) AS pct
+FROM
+  `httparchive.almanac.parsed_css`,
+  UNNEST(getKeyframePositions(css)) AS position
+JOIN (
   SELECT
-    client,
-    position,
-    COUNT(DISTINCT page) AS pages,
-    ANY_VALUE(total) AS total_pages,
-    COUNT(DISTINCT page) / ANY_VALUE(total) AS pct_pages,
-    COUNT(0) AS freq,
-    SUM(COUNT(0)) OVER (PARTITION BY client) AS total,
-    COUNT(0) / SUM(COUNT(0)) OVER (PARTITION BY client) AS pct
+    _TABLE_SUFFIX AS client,
+    COUNT(0) AS total
   FROM
-    `httparchive.almanac.parsed_css`,
-    UNNEST(getKeyframePositions(css)) AS position
-  JOIN (
-    SELECT
-      _TABLE_SUFFIX AS client,
-      COUNT(0) AS total
-    FROM
-      `httparchive.summary_pages.2021_07_01_*`
-    GROUP BY
-      client
-  )
-  USING (client)
-  WHERE
-    date = '2021-07-01'
+    `httparchive.summary_pages.2021_07_01_*`
   GROUP BY
-    client,
-    position
-  ORDER BY
-    pct DESC
+    client
 )
+USING (client)
+WHERE
+  date = '2021-07-01'
+GROUP BY
+  client,
+  position
+ORDER BY
+  pct DESC
 LIMIT 500

sql/2022/jamstack/jamstack_random_1000_urls.sql

@@ -5,4 +5,4 @@ FROM
 WHERE
   methodology = '2022' AND
   date = '2022-06-01'
-LIMIT 1000
+LIMIT 1000 -- noqa: AM09

sql/2024/jamstack/hugo_astro_next.sql

@@ -9,7 +9,7 @@ CREATE TEMPORARY FUNCTION GET_MAX_AGE(response_headers ARRAY<STRUCT<name STRING,
           UNNEST(response_headers) AS header
         WHERE
           LOWER(header.name) = 'cache-control'
-        LIMIT 1
+        LIMIT 1 -- noqa: AM09
       ),
       r'max-age=(\d+)'
     ) AS INT64

sql/2024/jamstack/jamstack-overview.sql

@@ -9,7 +9,7 @@ CREATE TEMPORARY FUNCTION GET_MAX_AGE(response_headers ARRAY<STRUCT<name STRING,
           UNNEST(response_headers) AS header
         WHERE
           LOWER(header.name) = 'cache-control'
-        LIMIT 1
+        LIMIT 1 -- noqa: AM09
       ),
       r'max-age=(\d+)'
     ) AS INT64

sql/2024/jamstack/jamstack_distribution_by_rank.sql

@@ -9,7 +9,7 @@ CREATE TEMPORARY FUNCTION GET_MAX_AGE(response_headers ARRAY<STRUCT<name STRING,
           UNNEST(response_headers) AS header
         WHERE
           LOWER(header.name) = 'cache-control'
-        LIMIT 1
+        LIMIT 1  -- noqa: AM09
       ),
       r'max-age=(\d+)'
     ) AS INT64

sql/2024/jamstack/js_frameworks.sql

@@ -9,7 +9,7 @@ CREATE TEMPORARY FUNCTION GET_MAX_AGE(response_headers ARRAY<STRUCT<name STRING,
           UNNEST(response_headers) AS header
         WHERE
           LOWER(header.name) = 'cache-control'
-        LIMIT 1
+        LIMIT 1 -- noqa: AM09
       ),
       r'max-age=(\d+)'
     ) AS INT64

sql/2024/jamstack/paas.sql

@@ -9,7 +9,7 @@ CREATE TEMPORARY FUNCTION GET_MAX_AGE(response_headers ARRAY<STRUCT<name STRING,
           UNNEST(response_headers) AS header
         WHERE
           LOWER(header.name) = 'cache-control'
-        LIMIT 1
+        LIMIT 1 -- noqa: AM09
       ),
       r'max-age=(\d+)'
     ) AS INT64

sql/2024/jamstack/ssg.sql

@@ -9,7 +9,7 @@ CREATE TEMPORARY FUNCTION GET_MAX_AGE(response_headers ARRAY<STRUCT<name STRING,
           UNNEST(response_headers) AS header
         WHERE
           LOWER(header.name) = 'cache-control'
-        LIMIT 1
+        LIMIT 1 -- noqa: AM09
       ),
       r'max-age=(\d+)'
     ) AS INT64

sql/2024/markup/content_encoding.sql

@@ -8,7 +8,7 @@ RETURNS STRING AS (
       UNNEST(response_headers) AS header
     WHERE
       LOWER(header.name) = 'content-encoding'
-    LIMIT 1
+    LIMIT 1 -- noqa: AM09
   )
 );
 

sql/2024/sustainability/video_autoplay_values.sql

@@ -8,7 +8,6 @@ WITH video_data AS (
   WHERE
     date = '2024-06-01' AND  -- Updated date
     is_root_page
-  LIMIT 10000  -- Limit the number of rows processed for faster testing
 )
 
 SELECT