feat: add write-to-files feature to deal with large change sets

This commit introduces a new input named 'write-to-files'. It enables writing
the lists of matching files to a corresponding file in addition to the output
'<filter-name>_files'. If set, the action will create the specified file with
the list of matching files. The file will be written in the format specified by
the `list-files` option and named after the filter. The path to the file will be
output as a variable named `<filter-name>_files_path`.

Author: diogokiss

README.md

@@ -142,6 +142,13 @@ For more information, see [CHANGELOG](https://github.com/dorny/paths-filter/blob
     # Default: none
     list-files: ''
 
+    # Enables writing the lists of matching files to a corresponding file.
+    # If set, the action will create the specified file with the list of matching files.
+    # The file will be written in the format specified by the `list-files` option and named
+    # after the filter. The path to the file will be relative to the working directory and
+    # exported as an output variable named `<filter-name>_files_path`.
+    write-to-files: ''
+
     # Relative path under $GITHUB_WORKSPACE where the repository was checked out.
     working-directory: ''
 
@@ -154,14 +161,14 @@ For more information, see [CHANGELOG](https://github.com/dorny/paths-filter/blob
     # Default: ${{ github.token }}
     token: ''
 
-    # Optional parameter to override the default behavior of file matching algorithm. 
+    # Optional parameter to override the default behavior of file matching algorithm.
     # By default files that match at least one pattern defined by the filters will be included.
     # This parameter allows to override the "at least one pattern" behavior to make it so that
-    # all of the patterns have to match or otherwise the file is excluded. 
-    # An example scenario where this is useful if you would like to match all 
-    # .ts files in a sub-directory but not .md files. 
-    # The filters below will match markdown files despite the exclusion syntax UNLESS 
-    # you specify 'every' as the predicate-quantifier parameter. When you do that, 
+    # all of the patterns have to match or otherwise the file is excluded.
+    # An example scenario where this is useful if you would like to match all
+    # .ts files in a sub-directory but not .md files.
+    # The filters below will match markdown files despite the exclusion syntax UNLESS
+    # you specify 'every' as the predicate-quantifier parameter. When you do that,
     # it will only match the .ts files in the subdirectory as expected.
     #
     # backend:
@@ -505,6 +512,37 @@ jobs:
 
 </details>
 
+<details>
+  <summary>Handle large change sets (2000+ files)</summary>
+
+```yaml
+- uses: dorny/paths-filter@v3
+  id: changed
+  with:
+    # Enable writing the files matching each filter to the disk in addition to the output '<filter_name>_files'.
+    # The path for each filter's file is output in the format '<filter_name>_files_path'.
+    write-to-files: true
+    list-files: json
+    filters: |
+      content:
+        - 'content/**'
+
+
+- name: List changed directories relative to the base directory
+  shell: bash
+  env:
+    BASE_DIR: ${{ inputs.base-directory }}
+    CHANGED_CONTENT_FILES_PATH: ${{ steps.changed.outputs.content_files_path }}
+  run: |
+    CHANGED_CONTENT_DIRECTORIES=$(cat "${CHANGED_CONTENT_FILES_PATH}" | xargs -n1 realpath -m --relative-to=${BASE_DIR} | cut -f1 -d / | sort -u)
+    for d in $CHANGED_CONTENT_DIRECTORIES
+    do
+      echo "Content directory change detected: ${d}"
+    done
+```
+
+</details>
+
 ### Custom processing of changed files
 
 <details>

action.yml

@@ -36,6 +36,13 @@ inputs:
                   Backslash escapes every potentially unsafe character.
     required: false
     default: none
+  write-to-files:
+    description: |
+      Enables writing the lists of matching files to a corresponding file in addition to the output '<filter-name>_files'.
+      If set, the action will create the specified file with the list of matching files.
+      The file will be written in the format specified by the `list-files` option and named
+      after the filter. The path to the file will be output as a variable named `<filter-name>_files_path`.
+    required: false
   initial-fetch-depth:
     description: |
       How many commits are initially fetched from base branch.

dist/index.js

@@ -552,15 +552,20 @@ var __importStar = (this && this.__importStar) || function (mod) {
     __setModuleDefault(result, mod);
     return result;
 };
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
 Object.defineProperty(exports, "__esModule", ({ value: true }));
 const fs = __importStar(__nccwpck_require__(7147));
 const core = __importStar(__nccwpck_require__(2186));
 const github = __importStar(__nccwpck_require__(5438));
+const path_1 = __importDefault(__nccwpck_require__(1017));
 const filter_1 = __nccwpck_require__(3707);
 const file_1 = __nccwpck_require__(4014);
 const git = __importStar(__nccwpck_require__(3374));
 const shell_escape_1 = __nccwpck_require__(4613);
 const csv_escape_1 = __nccwpck_require__(7402);
+const fs_1 = __nccwpck_require__(7147);
 async function run() {
     try {
         const workingDirectory = core.getInput('working-directory', { required: false });
@@ -573,6 +578,7 @@ async function run() {
         const filtersInput = core.getInput('filters', { required: true });
         const filtersYaml = isPathInput(filtersInput) ? getConfigFileContent(filtersInput) : filtersInput;
         const listFiles = core.getInput('list-files', { required: false }).toLowerCase() || 'none';
+        const writeToFiles = core.getInput('write-to-files', { required: false }).toLowerCase() === 'true';
         const initialFetchDepth = parseInt(core.getInput('initial-fetch-depth', { required: false })) || 10;
         const predicateQuantifier = core.getInput('predicate-quantifier', { required: false }) || filter_1.PredicateQuantifier.SOME;
         if (!isExportFormat(listFiles)) {
@@ -589,7 +595,7 @@ async function run() {
         const files = await getChangedFiles(token, base, ref, initialFetchDepth);
         core.info(`Detected ${files.length} changed files`);
         const results = filter.match(files);
-        exportResults(results, listFiles);
+        exportResults(results, listFiles, writeToFiles);
     }
     catch (error) {
         core.setFailed(getErrorMessage(error));
@@ -742,13 +748,14 @@ async function getChangedFilesFromApi(token, pullRequest) {
         core.endGroup();
     }
 }
-function exportResults(results, format) {
+function exportResults(results, format, writeToFiles) {
+    const tempDir = (0, fs_1.mkdtempSync)(path_1.default.join(process.cwd(), 'paths-filter-'));
     core.info('Results:');
     const changes = [];
     for (const [key, files] of Object.entries(results)) {
-        const value = files.length > 0;
-        core.startGroup(`Filter ${key} = ${value}`);
-        if (files.length > 0) {
+        const match = files.length > 0;
+        core.startGroup(`Filter ${key} = ${match}`);
+        if (match) {
             changes.push(key);
             core.info('Matching files:');
             for (const file of files) {
@@ -758,11 +765,18 @@ function exportResults(results, format) {
         else {
             core.info('Matching files: none');
         }
-        core.setOutput(key, value);
+        core.setOutput(key, match);
         core.setOutput(`${key}_count`, files.length);
         if (format !== 'none') {
             const filesValue = serializeExport(files, format);
             core.setOutput(`${key}_files`, filesValue);
+            if (writeToFiles) {
+                const ext = format === 'json' ? 'json' : 'txt';
+                const filePath = path_1.default.join(tempDir, `${key}-files.${ext}`);
+                fs.writeFileSync(filePath, filesValue);
+                core.info(`Matching files list for filter '${key}' written to '${filePath}'`);
+                core.setOutput(`${key}_files_path`, filePath);
+            }
         }
         core.endGroup();
     }

src/main.ts

@@ -1,6 +1,7 @@
 import * as fs from 'fs'
 import * as core from '@actions/core'
 import * as github from '@actions/github'
+import path from 'path'
 import {GetResponseDataTypeFromEndpointMethod} from '@octokit/types'
 import {PushEvent, PullRequestEvent} from '@octokit/webhooks-types'
 
@@ -16,6 +17,7 @@ import {File, ChangeStatus} from './file'
 import * as git from './git'
 import {backslashEscape, shellEscape} from './list-format/shell-escape'
 import {csvEscape} from './list-format/csv-escape'
+import {mkdtempSync} from 'fs'
 
 type ExportFormat = 'none' | 'csv' | 'json' | 'shell' | 'escape'
 
@@ -32,6 +34,7 @@ async function run(): Promise<void> {
     const filtersInput = core.getInput('filters', {required: true})
     const filtersYaml = isPathInput(filtersInput) ? getConfigFileContent(filtersInput) : filtersInput
     const listFiles = core.getInput('list-files', {required: false}).toLowerCase() || 'none'
+    const writeToFiles = core.getInput('write-to-files', {required: false}).toLowerCase() === 'true'
     const initialFetchDepth = parseInt(core.getInput('initial-fetch-depth', {required: false})) || 10
     const predicateQuantifier = core.getInput('predicate-quantifier', {required: false}) || PredicateQuantifier.SOME
 
@@ -52,7 +55,7 @@ async function run(): Promise<void> {
     const files = await getChangedFiles(token, base, ref, initialFetchDepth)
     core.info(`Detected ${files.length} changed files`)
     const results = filter.match(files)
-    exportResults(results, listFiles)
+    exportResults(results, listFiles, writeToFiles)
   } catch (error) {
     core.setFailed(getErrorMessage(error))
   }
@@ -228,13 +231,15 @@ async function getChangedFilesFromApi(token: string, pullRequest: PullRequestEve
   }
 }
 
-function exportResults(results: FilterResults, format: ExportFormat): void {
+function exportResults(results: FilterResults, format: ExportFormat, writeToFiles: boolean): void {
+  const tempDir = mkdtempSync(path.join(process.cwd(), 'paths-filter-'))
+
   core.info('Results:')
   const changes = []
   for (const [key, files] of Object.entries(results)) {
-    const value = files.length > 0
-    core.startGroup(`Filter ${key} = ${value}`)
-    if (files.length > 0) {
+    const match = files.length > 0
+    core.startGroup(`Filter ${key} = ${match}`)
+    if (match) {
       changes.push(key)
       core.info('Matching files:')
       for (const file of files) {
@@ -244,12 +249,21 @@ function exportResults(results: FilterResults, format: ExportFormat): void {
       core.info('Matching files: none')
     }
 
-    core.setOutput(key, value)
+    core.setOutput(key, match)
     core.setOutput(`${key}_count`, files.length)
     if (format !== 'none') {
       const filesValue = serializeExport(files, format)
       core.setOutput(`${key}_files`, filesValue)
+
+      if (writeToFiles) {
+        const ext = format === 'json' ? 'json' : 'txt'
+        const filePath = path.join(tempDir, `${key}-files.${ext}`)
+        fs.writeFileSync(filePath, filesValue)
+        core.info(`Matching files list for filter '${key}' written to '${filePath}'`)
+        core.setOutput(`${key}_files_path`, filePath)
+      }
     }
+
     core.endGroup()
   }