This is an automated email from the ASF dual-hosted git repository. dsmiley pushed a commit to branch branch_10x in repository https://gitbox.apache.org/repos/asf/solr.git
commit 93ccce22a8ec5b5a1e4dac6ab5ed33dc11c8e507 Author: sarkaramrit1993 <[email protected]> AuthorDate: Wed Oct 7 13:01:52 2026 -0700 SOLR-18031: Respect --filetypes for files and globs in bin/solr post (#5032) (cherry picked from commit d97bc70cefb84614ed6ecb61792df1155817340d) --- changelog/unreleased/solr-18031-post-filetypes.yml | 7 ++++ .../src/java/org/apache/solr/cli/PostTool.java | 19 +++++++--- .../src/test/org/apache/solr/cli/PostToolTest.java | 40 ++++++++++++++++++++++ .../modules/indexing-guide/pages/post-command.adoc | 3 ++ 4 files changed, 65 insertions(+), 4 deletions(-) diff --git a/changelog/unreleased/solr-18031-post-filetypes.yml b/changelog/unreleased/solr-18031-post-filetypes.yml new file mode 100644 index 00000000000..65e19647a09 --- /dev/null +++ b/changelog/unreleased/solr-18031-post-filetypes.yml @@ -0,0 +1,7 @@ +title: bin/solr post now honors --filetypes for files and globs, and reports the right file count. +type: fixed +authors: + - name: Amrit Sarkar +links: + - name: SOLR-18031 + url: https://issues.apache.org/jira/browse/SOLR-18031 diff --git a/solr/core/src/java/org/apache/solr/cli/PostTool.java b/solr/core/src/java/org/apache/solr/cli/PostTool.java index 8c455cefa3a..afd1c50467a 100644 --- a/solr/core/src/java/org/apache/solr/cli/PostTool.java +++ b/solr/core/src/java/org/apache/solr/cli/PostTool.java @@ -482,6 +482,10 @@ public class PostTool extends ToolBase { return Files.exists(srcFile); } + private boolean matchesFileTypes(Path file) { + return !auto || fileFilter == null || fileFilter.test(file); + } + private static Collection<Path> listFiles(Path directory, Predicate<Path> fileFilter) throws IOException { Predicate<Path> filter = fileFilter != null ? fileFilter : p -> true; @@ -523,7 +527,7 @@ public class PostTool extends ToolBase { reset(); int filesPosted = 0; for (int j = startIndexInArgs; j < args.length; j++) { - filesPosted = getFilesPosted(out, type, args[j]); + filesPosted += getFilesPosted(out, type, args[j]); } return filesPosted; } @@ -536,7 +540,11 @@ public class PostTool extends ToolBase { if (isValidPath && Files.isDirectory(srcFile) && Files.isReadable(srcFile)) { filesPosted += postDirectory(srcFile, out, type); } else if (isValidPath && Files.isRegularFile(srcFile) && Files.isReadable(srcFile)) { - filesPosted += postFiles(List.of(srcFile), out, type); + if (matchesFileTypes(srcFile)) { + filesPosted += postFiles(List.of(srcFile), out, type); + } else { + warn("Skipping " + srcFile + ", its file type is not in " + fileTypes); + } } else { filesPosted += handleGlob(src, out, type); } @@ -612,9 +620,12 @@ public class PostTool extends ToolBase { } String fileGlob = globPath.getFileName().toString(); GlobFilter ff = new GlobFilter(fileGlob, false); - Collection<Path> fileList = listFiles(parent, ff); - if (fileList.isEmpty()) { + Collection<Path> matches = listFiles(parent, ff); + List<Path> fileList = matches.stream().filter(this::matchesFileTypes).toList(); + if (matches.isEmpty()) { warn("No files or directories matching " + globPath); + } else if (fileList.isEmpty()) { + warn("No files matching " + globPath + " have a file type in " + fileTypes); } else { filesPosted = postFiles(fileList, out, type); } diff --git a/solr/core/src/test/org/apache/solr/cli/PostToolTest.java b/solr/core/src/test/org/apache/solr/cli/PostToolTest.java index a238bd62c15..5390fbeafcf 100644 --- a/solr/core/src/test/org/apache/solr/cli/PostToolTest.java +++ b/solr/core/src/test/org/apache/solr/cli/PostToolTest.java @@ -33,6 +33,7 @@ import java.util.HashSet; import java.util.List; import java.util.Map; import java.util.Set; +import org.apache.lucene.util.Constants; import org.apache.solr.SolrTestCaseJ4; import org.apache.solr.client.solrj.SolrRequest; import org.apache.solr.client.solrj.SolrResponse; @@ -293,6 +294,45 @@ public class PostToolTest extends SolrCloudTestCase { assertEquals(2, num); } + @Test + public void testFileTypesAppliedToFileArgs() throws IOException { + Path dir = createTempDir(); + Path a = Files.writeString(dir.resolve("a.xml"), "<add/>"); + Path b = Files.writeString(dir.resolve("B.XML"), "<add/>"); + Path c = Files.writeString(dir.resolve("c.csv"), "id\n1\n"); + Path d = Files.writeString(dir.resolve("noext"), "text"); + + PostTool postTool = dryRunPostTool("xml"); + String[] files = {a.toString(), b.toString(), c.toString(), d.toString()}; + assertEquals(2, postTool.postFiles(files, 0, null, null)); + + postTool.auto = false; + assertEquals(4, postTool.postFiles(files, 0, null, null)); + } + + @Test + public void testFileTypesAppliedToGlobs() throws IOException { + assumeFalse("'*' is not a valid path character on Windows", Constants.WINDOWS); + Path dir = createTempDir(); + Files.writeString(dir.resolve("a.xml"), "<add/>"); + Files.writeString(dir.resolve("B.XML"), "<add/>"); + Files.writeString(dir.resolve("c.csv"), "id\n1\n"); + + PostTool postTool = dryRunPostTool("xml"); + assertEquals(2, postTool.postFiles(new String[] {dir + "/*"}, 0, null, null)); + assertEquals(0, postTool.postFiles(new String[] {dir + "/*.csv"}, 0, null, null)); + } + + private PostTool dryRunPostTool(String fileTypes) { + PostTool postTool = new PostTool(new CLITestHelper.TestingRuntime(false)); + postTool.recursive = 0; + postTool.dryRun = true; + postTool.solrUpdateUrl = URI.create("http://localhost:8983/solr/fake/update";); + postTool.fileTypes = fileTypes; + postTool.fileFilter = postTool.getFileFilterFromFileTypes(fileTypes); + return postTool; + } + @Test public void testDetectingIfRecursionPossibleInFilesMode() throws IOException { ToolRuntime runtime = new CLITestHelper.TestingRuntime(false); diff --git a/solr/solr-ref-guide/modules/indexing-guide/pages/post-command.adoc b/solr/solr-ref-guide/modules/indexing-guide/pages/post-command.adoc index 9c008b51ab3..b7a8cfb8738 100644 --- a/solr/solr-ref-guide/modules/indexing-guide/pages/post-command.adoc +++ b/solr/solr-ref-guide/modules/indexing-guide/pages/post-command.adoc @@ -121,6 +121,9 @@ Index all CSV and JSON files into `gettingstarted` from current directory: $ bin/solr post -c gettingstarted --filetypes json,csv . ---- +`--filetypes` also applies to files named on the command line, including a shell-expanded `*`. +A file with any other extension is skipped with a warning, unless `--type` is set. + Index a tab-separated file into `gettingstarted`: [,console]
