Change HtmlSanityCheckTask.sourceDocuments to a org.gradle.api.file.F…

…ileCollection to benefit from finding and filtering capabilities of the FileCollection API. (#267)

Fixes #264.

README.adoc

@@ -74,8 +74,8 @@ This task exposes a few properties as part of its configuration:
 [horizontal]
 sourceDir:: (mandatory) directory where the html files are located. Type: File. Default: `build/docs`.
 sourceDocuments:: (optional) an override to process several source files, which may be a subset of all
-                      files available in [x-]`${sourceDir}`. Type: Set.
-                      Defaults to all files in [x-]`${sourceDir}`.
+                      files available in [x-]`${sourceDir}`. Type: `org.gradle.api.file.FileCollection`.
+                      Defaults to all files in [x-]`${sourceDir}` whose names end with `.html`.
 
 checkingResultsDir:: (optional) directory where the checking results written to.
                       Defaults to `${buildDir}/report/htmlchecks/`
@@ -101,9 +101,6 @@ apply plugin: 'org.aim42.htmlSanityCheck'
 htmlSanityCheck {
     sourceDir = new File( "$buildDir/docs" )
 
-    // files to check - in Set-notation
-    sourceDocuments = [ "one-file.html", "another-file.html", "index.html"]
-
     // where to put results of sanityChecks...
     checkingResultsDir = new File( "$buildDir/report/htmlchecks" )
 
@@ -188,8 +185,10 @@ htmlSanityCheck {
 
     sourceDir = new File( htmlOutputPath )
 
-    // files to check, in Set-notation
-    sourceDocuments = [ "many-errors.html", "no-errors.html"]
+    // files to check, specified as a file tree with filtering
+    sourceDocuments = fileTree(sourceDir) {
+        include "many-errors.html", "no-errors.html"
+    }
 
     // where to put results of sanityChecks...
     checkingResultsDir = new File( checkingResultsPath )

src/main/groovy/org/aim42/filesystem/FileCollector.groovy

@@ -10,34 +10,8 @@ import java.util.regex.Pattern
 
 class FileCollector {
 
-    // (?i): ignore-case
-    // htm(l): either htm or html
-    // at least one character left of the dot
-    public final static Pattern HTML_FILE_EXTENSION_PATTERN = ~/(?i).+\.htm(l)?$/
-
-    public final static Closure<Boolean> INCLUDE_HTML_FILES = { File file ->
-        // allow only html or htm files
-        (file.name ==~ HTML_FILE_EXTENSION_PATTERN)
-    }
-
     public final static String IMAGE_FILE_EXTENSION_PATTERN = ~/(?i).+\.(jpg|jpeg|png|gif|bmp|svg)?$/
 
-
-    public static Boolean isHtmlFile(File file) {
-        return (isHtmlFile(file.absolutePath)
-                && file.isFile())
-    }
-
-    /**
-     * checks if @param fileName represents a valid html file,
-     * ignoring the files' content!
-     * @param fileName
-     * @return
-     */
-    public static Boolean isHtmlFile(String fileName) {
-        return (fileName ==~ HTML_FILE_EXTENSION_PATTERN)
-    }
-
     /**
      * checks if @param fileName represents an image file name,
      * ignoring the files' content.
@@ -47,37 +21,6 @@ class FileCollector {
         return (fileName ==~ IMAGE_FILE_EXTENSION_PATTERN)
     }
 
-    /**
-     * The main use-case for this class: returns all configured html files as Set<File>.
-     * @param srcDir (mandatory) Directory where the html files are located. Type: File. Default: build/docs
-     *            can be given as File or String
-     * @param sourceDocs (optional) Defaults to all files in ${srcDir}.
-     */
-    public static Set<File> getHtmlFilesToCheck(File srcDir, Collection<String> sourceDocs) {
-        // first case: no document names given -> return all html files
-        // in directory
-        if ((sourceDocs == null) || (sourceDocs?.empty)) {
-            return getAllHtmlFilesFromDirectory(srcDir)
-        } else {
-            return getAllConfiguredHtmlFiles(srcDir, sourceDocs)
-        }
-    }
-
-    // if for other types of input parameters, convert and delegate
-
-    public static Set<File> getHtmlFilesToCheck(File srcDir, String sourceDoc) {
-        return getHtmlFilesToCheck(srcDir, [sourceDoc])
-    }
-
-    public static Set<File> getHtmlFilesToCheck(String srcDirName, ArrayList<String> sourceDocs) {
-        return getHtmlFilesToCheck(new File(srcDirName), sourceDocs)
-    }
-
-    public static Set<File> getHtmlFilesToCheck(String srcDirName, String sourceDocName) {
-        return getHtmlFilesToCheck(new File(srcDirName), [sourceDocName])
-    }
-
-
     /**
      * returns all configured image files as Set<File>
      *
@@ -94,25 +37,6 @@ class FileCollector {
         }
     }
 
-    /**
-     * returns all html files in a given directory.
-     * (recursively looks in subdirectories)
-     * @param dir where to look for matching files
-     * @return all files with appropriate extension
-     */
-    public static Set<File> getAllHtmlFilesFromDirectory(File dir) {
-        Set<File> files = new HashSet<File>()
-
-        // scan only files, not directories
-        dir.eachFileRecurse(FileType.FILES) { file ->
-            if (isHtmlFile(file)) {
-                files.add(file)
-            }
-        }
-
-        return files
-    }
-
     /**
      * returns all image files in a given directory.
      * (recursively looks in subdirectories)
@@ -132,24 +56,6 @@ class FileCollector {
         return files
     }
 
-    /**
-     * returns all configured html files from @param srcDocs
-     * which really exist below @param srcDocs
-     */
-    public static Set<File> getAllConfiguredHtmlFiles(File srcDir, Collection<String> srcDocs) {
-        Set<File> files = new HashSet<File>()
-
-        srcDocs.each { configuredFileName ->
-            File file = new File(srcDir, configuredFileName)
-            if (file.exists()) {
-                files.add(file)
-            }
-        }
-
-        return files
-
-    }
-
     /**
      * returns all configured image files from @param srcDocs
      * which really exist below @param srcDocs

src/main/groovy/org/aim42/htmlsanitycheck/AllChecksRunner.groovy

@@ -68,11 +68,7 @@ class AllChecksRunner {
 
         myConfig = pConfig
 
-        this.filesToCheck = FileCollector.getHtmlFilesToCheck(
-                myConfig.getConfigItemByName(Configuration.ITEM_NAME_sourceDir),
-                myConfig.getConfigItemByName (Configuration.ITEM_NAME_sourceDocuments)
-        )
-
+        this.filesToCheck = myConfig.getConfigItemByName(Configuration.ITEM_NAME_sourceDocuments)
 
         // TODO: #185 (checker classes shall be detected automatically (aka CheckerFactory)
         // CheckerFactory needs the configuration

src/main/groovy/org/aim42/htmlsanitycheck/Configuration.groovy

@@ -113,7 +113,7 @@ class Configuration {
     /**
      * convenience method for simplified testing
      */
-     synchronized void addSourceFileConfiguration(File srcDir, Set<String> srcDocs) {
+     synchronized void addSourceFileConfiguration(File srcDir, Collection<File> srcDocs) {
         addConfigurationItem(ITEM_NAME_sourceDir, srcDir)
         addConfigurationItem(ITEM_NAME_sourceDocuments, srcDocs)
     }
@@ -232,34 +232,18 @@ class Configuration {
             throw new MisconfigurationException("source directory must not be null")
         }
 
-        // cannot check if both input params are null
-        if ((srcDir == null) && (srcDocs == null)) {
-            throw new MisconfigurationException("both sourceDir and sourceDocs were null")
-        }
-
-        // no srcDir was given and empty SrcDocs
-        if ((!srcDir) && (srcDocs != null)) {
-            if ((srcDocs?.empty)) {
-                throw new MisconfigurationException("both sourceDir and sourceDocs must not be empty")
-            }
-        }
-        // non-existing srcDir is absurd too
         if ((!srcDir.exists())) {
             throw new MisconfigurationException("given sourceDir $srcDir does not exist.")
         }
 
-        // if srcDir exists but is empty... no good :-(
-        if ((srcDir.exists())
-                && (srcDir.isDirectory())
-                && (srcDir.directorySize() == 0)) {
-            throw new MisconfigurationException("given sourceDir $srcDir is empty")
+        // cannot check if both input params are null
+        if (srcDocs == null) {
+            throw new MisconfigurationException("source documents must not be null")
         }
 
-        // if srcDir exists but does not contain any html file... no good
-        if ((srcDir.exists())
-                && (srcDir.isDirectory())
-                && (FileCollector.getAllHtmlFilesFromDirectory(srcDir).size() == 0)) {
-            throw new MisconfigurationException("no html file found in $srcDir")
+        // empty SrcDocs
+        if (srcDocs.empty) {
+            throw new MisconfigurationException("source documents must not be empty")
         }
 
         Object checksToExecute = getConfigItemByName(Configuration.ITEM_NAME_checksToExecute)

src/main/groovy/org/aim42/htmlsanitycheck/HtmlSanityCheckTask.groovy

@@ -2,11 +2,12 @@ package org.aim42.htmlsanitycheck
 
 import org.aim42.htmlsanitycheck.check.AllCheckers
 import org.gradle.api.DefaultTask
-
-// see end-of-file for license information
 import org.gradle.api.GradleException
+import org.gradle.api.file.FileCollection
 import org.gradle.api.tasks.*
 
+// see end-of-file for license information
+
 /**
  * Entry class for the gradle-plugin.
  * Handles parameter-passing from gradle build scripts,
@@ -19,9 +20,8 @@ class HtmlSanityCheckTask extends DefaultTask {
 
     //
     // we support checking several named files
-    @Optional
-    @Input
-    Set<String> sourceDocuments
+    @InputFiles
+    FileCollection sourceDocuments
 
     // or all (html) files in a directory
     @InputDirectory
@@ -99,10 +99,18 @@ class HtmlSanityCheckTask extends DefaultTask {
 
     }
 
-    /**
-     * entry point for several html sanity checks
-     * @author Gernot Starke <[email protected]>
-     */
+    void setSourceDir(File sourceDir) {
+        this.sourceDir = sourceDir
+        if (sourceDocuments == null) {
+            sourceDocuments = project.fileTree(sourceDir)
+            sourceDocuments.include('**/*.html')
+        }
+    }
+
+/**
+ * entry point for several html sanity checks
+ * @author Gernot Starke <[email protected]>
+ */
     @TaskAction
     public void sanityCheckHtml() {
 
@@ -131,7 +139,7 @@ class HtmlSanityCheckTask extends DefaultTask {
             logger.info("allFilesToCheck" + allFilesToCheck.toString(), "")
 
             // create an AllChecksRunner...
-            def allChecksRunner = new AllChecksRunner( myConfig )
+            def allChecksRunner = new AllChecksRunner(myConfig)
 
             // ... and perform the actual checks
             def allChecks = allChecksRunner.performAllChecks()
@@ -166,7 +174,7 @@ See ${checkingResultsDir} for a detailed report."""
         Configuration tmpConfig = new Configuration()
 
         tmpConfig.with {
-            addConfigurationItem(Configuration.ITEM_NAME_sourceDocuments, sourceDocuments)
+            addConfigurationItem(Configuration.ITEM_NAME_sourceDocuments, sourceDocuments.files)
             addConfigurationItem(Configuration.ITEM_NAME_sourceDir, sourceDir)
             addConfigurationItem(Configuration.ITEM_NAME_checkingResultsDir, checkingResultsDir)
             addConfigurationItem(Configuration.ITEM_NAME_junitResultsDir, junitResultsDir)