[
https://issues.apache.org/jira/browse/TIKA-4856?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18110729#comment-18110729
]
ASF GitHub Bot commented on TIKA-4856:
--------------------------------------
dschmidt commented on code in PR #3096:
URL: https://github.com/apache/tika/pull/3096#discussion_r3916005607
##########
tika-server/tika-server-core/src/main/java/org/apache/tika/server/core/resource/ThumbnailDefaults.java:
##########
@@ -0,0 +1,198 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.tika.server.core.resource;
+
+import java.io.IOException;
+import java.io.UncheckedIOException;
+import java.util.Iterator;
+import java.util.LinkedHashMap;
+import java.util.Map;
+
+import com.fasterxml.jackson.databind.JsonNode;
+import com.fasterxml.jackson.databind.ObjectMapper;
+import com.fasterxml.jackson.databind.node.ObjectNode;
+
+import org.apache.tika.config.loader.TikaJsonConfig;
+import org.apache.tika.parser.ParseContext;
+
+/**
+ * The parser configuration that makes a parse yield the document thumbnail
+ * as a raster image: the first PDF page rendered, the EMF/WMF thumbnail of
+ * an Office document rendered (that one only, not the pictures of embedded
+ * objects), in colour: the renderer's default is the grayscale OCR wants.
+ * The stored thumbnails of the other formats need no configuration.
+ * <p>
+ * Applied by {@code renderThumbnails=true} on {@code /rmeta}, {@code /unpack}
+ * and {@code /unpack/all}, and by {@code /unpack/thumbnail}. Three layers,
+ * each overriding the one before: the built-in defaults below, a
+ * {@code thumbnail-defaults} block in the server config with the same shape
+ * as a request config (parser configurations keyed by component name), and
+ * the request's own config part.
+ * <pre>
+ * "thumbnail-defaults": {
+ * "pdf-parser": {"imageStrategy": "RENDER_PAGES_AT_PAGE_END",
"maxRenderedPages": 1,
+ * "ocr": {"dpi": 150}}
+ * }
+ * </pre>
+ * A configuration for a parser that is not installed is never read, so the
+ * defaults are harmless on a server without that parser.
+ */
+public final class ThumbnailDefaults {
+
+ public static final String CONFIG_KEY = "thumbnail-defaults";
+
+ private static final ObjectMapper MAPPER = new ObjectMapper();
+
+ private static final String BUILT_IN = """
+ {
+ "pdf-parser": {
+ "imageStrategy": "RENDER_PAGES_AT_PAGE_END",
+ "maxRenderedPages": 1,
+ "ocr": {"dpi": 96, "imageType": "RGB"}
+ },
+ "emf-parser": {"renderImage": true,
"renderOnlyEmbeddedResourceTypes": ["THUMBNAIL"]},
+ "wmf-parser": {"renderImage": true,
"renderOnlyEmbeddedResourceTypes": ["THUMBNAIL"]}
+ }
+ """;
+
+ /**
+ * Parser configurations keyed by component name, in application order.
+ */
+ private final Map<String, ObjectNode> components;
+
+ private ThumbnailDefaults(Map<String, ObjectNode> components) {
+ this.components = components;
+ }
+
+ /**
+ * No defaults at all, a base to {@link #with(String)} settings on.
+ */
+ public static ThumbnailDefaults none() {
+ return new ThumbnailDefaults(new LinkedHashMap<>());
+ }
+
+ /**
+ * These defaults with another set merged in, field by field.
+ */
+ public ThumbnailDefaults with(ThumbnailDefaults other) {
+ ThumbnailDefaults merged = this;
+ for (Map.Entry<String, ObjectNode> component :
other.components.entrySet()) {
+ merged = merged.with("{\"" + component.getKey() + "\": " +
component.getValue() + "}");
+ }
+ return merged;
Review Comment:
Fixed in 02d86467e1: both `with` overloads merge the nodes now, and a
component only the other side has is copied rather than shared between the two
sets of defaults.
##########
tika-server/tika-server-core/src/main/java/org/apache/tika/server/core/resource/UnpackerResource.java:
##########
@@ -215,15 +273,184 @@ public Response unpackAll(InputStream is, @Context
HttpHeaders httpHeaders, @Con
@POST
@Consumes("multipart/form-data")
@Produces("application/zip")
- public Response unpackAllWithConfig(List<Attachment> attachments, @Context
HttpHeaders httpHeaders, @Context UriInfo info) throws Exception {
+ public Response unpackAllWithConfig(List<Attachment> attachments, @Context
HttpHeaders httpHeaders, @Context UriInfo info,
+ @QueryParam("renderThumbnails") boolean
renderThumbnails) throws Exception {
ParseContext pc = tikaResource.createRequestContext();
Metadata metadata = tikaResource.newRequestMetadata();
try (TikaInputStream tis =
tikaResource.setupMultipartConfig(attachments, metadata, pc)) {
TikaResource.logRequest(LOG, "/unpack/all", metadata);
+ if (renderThumbnails) {
+ //under the request's config, which setupMultipartConfig has
already merged
+ tikaResource.getThumbnailDefaults().applyTo(pc);
+ }
return doUnpack(tis, metadata, pc, true);
}
}
+ /**
+ * Returns the document thumbnail with its metadata (simple PUT).
+ */
+ @jakarta.ws.rs.Path("/thumbnail")
+ @PUT
+ @Produces("application/json")
+ public Response unpackThumbnail(InputStream is, @Context HttpHeaders
httpHeaders,
+ @QueryParam("renderThumbnails") boolean
renderThumbnails) throws Exception {
+ ParseContext pc = tikaResource.createRequestContext();
+ Metadata metadata = tikaResource.newRequestMetadata();
+ try (TikaInputStream tis = TikaInputStream.get(is)) {
+ fillMetadata(null, metadata, httpHeaders.getRequestHeaders());
+ TikaResource.logRequest(LOG, "/unpack/thumbnail", metadata);
+ return doUnpackThumbnail(tis, metadata, pc, renderThumbnails);
+ }
+ }
+
+ /**
+ * Returns the document thumbnail with its metadata (multipart POST,
"file" part).
+ */
+ @jakarta.ws.rs.Path("/thumbnail")
+ @POST
+ @Consumes("multipart/form-data")
+ @Produces("application/json")
+ public Response unpackThumbnailMultipart(List<Attachment> attachments,
@Context HttpHeaders httpHeaders,
+ @QueryParam("renderThumbnails")
boolean renderThumbnails)
+ throws Exception {
+ ParseContext pc = tikaResource.createRequestContext();
+ Metadata metadata = tikaResource.newRequestMetadata();
+ try (TikaInputStream tis =
tikaResource.setupMultipartConfig(attachments, metadata, pc)) {
+ TikaResource.logRequest(LOG, "/unpack/thumbnail", metadata);
+ return doUnpackThumbnail(tis, metadata, pc, renderThumbnails);
+ }
+ }
+
+ private static final ObjectMapper MAPPER = new ObjectMapper();
+ private static final String METADATA_SUFFIX = ".metadata.json";
+ /**
+ * A thumbnail travels base64-encoded inside a JSON object, so it is
+ * bounded here regardless of the unpack limits; camera previews and
+ * page renderings are a few MB at most.
+ */
+ static final long MAX_THUMBNAIL_BYTES = 32L * 1024 * 1024;
+ /**
+ * What {@code /unpack/thumbnail} adds regardless of rendering: the text
+ * of the images is not wanted.
+ */
+ private static final ThumbnailDefaults NO_OCR = ThumbnailDefaults.none()
+ .with("{\"pdf-parser\": {\"ocr\": {\"strategy\": \"NO_OCR\"}}, "
+ + "\"tesseract-ocr-parser\": {\"skipOcr\": true}}");
+
+ /**
+ * Parses in unpack mode with the thumbnail configuration, then selects
+ * the thumbnail among the extracted embedded documents.
+ */
+ private Response doUnpackThumbnail(TikaInputStream tis, Metadata metadata,
ParseContext pc,
+ boolean renderThumbnails) throws
Exception {
+ PipesParsingHelper helper = tikaResource.getPipesParsingHelper();
+ if (helper == null) {
+ throw new WebApplicationException("Pipes-based parsing is not
enabled", Response.Status.SERVICE_UNAVAILABLE);
+ }
+ configureThumbnailParse(pc, renderThumbnails);
+
+ PipesParsingHelper.UnpackResult result = helper.parseUnpack(tis,
metadata, pc, false);
+ if (result.zipFile() == null) {
+ throw new WebApplicationException(Response.Status.NO_CONTENT);
+ }
+ try (ZipFile zip = new ZipFile(result.zipFile().toFile())) {
+ Map<String, Metadata> extracted = readExtractedMetadata(zip);
+ Metadata thumbnail = ThumbnailSelector.select(new
ArrayList<>(extracted.values()));
+ if (thumbnail == null) {
+ throw new WebApplicationException(Response.Status.NO_CONTENT);
+ }
+ String entryName = null;
+ for (Map.Entry<String, Metadata> e : extracted.entrySet()) {
+ if (e.getValue() == thumbnail) {
+ entryName = e.getKey();
Review Comment:
The `==` is deliberate, the selector hands back one of the very objects in
that map, but you are right that it kept scanning after finding it. Fixed in
02d86467e1 with a break and a line saying why identity is the right test.
##########
tika-server/tika-server-core/src/main/java/org/apache/tika/server/core/resource/UnpackerResource.java:
##########
@@ -215,15 +273,184 @@ public Response unpackAll(InputStream is, @Context
HttpHeaders httpHeaders, @Con
@POST
@Consumes("multipart/form-data")
@Produces("application/zip")
- public Response unpackAllWithConfig(List<Attachment> attachments, @Context
HttpHeaders httpHeaders, @Context UriInfo info) throws Exception {
+ public Response unpackAllWithConfig(List<Attachment> attachments, @Context
HttpHeaders httpHeaders, @Context UriInfo info,
+ @QueryParam("renderThumbnails") boolean
renderThumbnails) throws Exception {
ParseContext pc = tikaResource.createRequestContext();
Metadata metadata = tikaResource.newRequestMetadata();
try (TikaInputStream tis =
tikaResource.setupMultipartConfig(attachments, metadata, pc)) {
TikaResource.logRequest(LOG, "/unpack/all", metadata);
+ if (renderThumbnails) {
+ //under the request's config, which setupMultipartConfig has
already merged
+ tikaResource.getThumbnailDefaults().applyTo(pc);
+ }
return doUnpack(tis, metadata, pc, true);
}
}
+ /**
+ * Returns the document thumbnail with its metadata (simple PUT).
+ */
+ @jakarta.ws.rs.Path("/thumbnail")
+ @PUT
+ @Produces("application/json")
+ public Response unpackThumbnail(InputStream is, @Context HttpHeaders
httpHeaders,
+ @QueryParam("renderThumbnails") boolean
renderThumbnails) throws Exception {
+ ParseContext pc = tikaResource.createRequestContext();
+ Metadata metadata = tikaResource.newRequestMetadata();
+ try (TikaInputStream tis = TikaInputStream.get(is)) {
+ fillMetadata(null, metadata, httpHeaders.getRequestHeaders());
+ TikaResource.logRequest(LOG, "/unpack/thumbnail", metadata);
+ return doUnpackThumbnail(tis, metadata, pc, renderThumbnails);
+ }
+ }
+
+ /**
+ * Returns the document thumbnail with its metadata (multipart POST,
"file" part).
+ */
+ @jakarta.ws.rs.Path("/thumbnail")
+ @POST
+ @Consumes("multipart/form-data")
+ @Produces("application/json")
+ public Response unpackThumbnailMultipart(List<Attachment> attachments,
@Context HttpHeaders httpHeaders,
+ @QueryParam("renderThumbnails")
boolean renderThumbnails)
+ throws Exception {
+ ParseContext pc = tikaResource.createRequestContext();
+ Metadata metadata = tikaResource.newRequestMetadata();
+ try (TikaInputStream tis =
tikaResource.setupMultipartConfig(attachments, metadata, pc)) {
+ TikaResource.logRequest(LOG, "/unpack/thumbnail", metadata);
+ return doUnpackThumbnail(tis, metadata, pc, renderThumbnails);
+ }
+ }
+
+ private static final ObjectMapper MAPPER = new ObjectMapper();
+ private static final String METADATA_SUFFIX = ".metadata.json";
+ /**
+ * A thumbnail travels base64-encoded inside a JSON object, so it is
+ * bounded here regardless of the unpack limits; camera previews and
+ * page renderings are a few MB at most.
+ */
+ static final long MAX_THUMBNAIL_BYTES = 32L * 1024 * 1024;
+ /**
+ * What {@code /unpack/thumbnail} adds regardless of rendering: the text
+ * of the images is not wanted.
+ */
+ private static final ThumbnailDefaults NO_OCR = ThumbnailDefaults.none()
+ .with("{\"pdf-parser\": {\"ocr\": {\"strategy\": \"NO_OCR\"}}, "
+ + "\"tesseract-ocr-parser\": {\"skipOcr\": true}}");
+
+ /**
+ * Parses in unpack mode with the thumbnail configuration, then selects
+ * the thumbnail among the extracted embedded documents.
+ */
+ private Response doUnpackThumbnail(TikaInputStream tis, Metadata metadata,
ParseContext pc,
+ boolean renderThumbnails) throws
Exception {
+ PipesParsingHelper helper = tikaResource.getPipesParsingHelper();
+ if (helper == null) {
+ throw new WebApplicationException("Pipes-based parsing is not
enabled", Response.Status.SERVICE_UNAVAILABLE);
+ }
+ configureThumbnailParse(pc, renderThumbnails);
+
+ PipesParsingHelper.UnpackResult result = helper.parseUnpack(tis,
metadata, pc, false);
+ if (result.zipFile() == null) {
+ throw new WebApplicationException(Response.Status.NO_CONTENT);
+ }
+ try (ZipFile zip = new ZipFile(result.zipFile().toFile())) {
+ Map<String, Metadata> extracted = readExtractedMetadata(zip);
+ Metadata thumbnail = ThumbnailSelector.select(new
ArrayList<>(extracted.values()));
+ if (thumbnail == null) {
+ throw new WebApplicationException(Response.Status.NO_CONTENT);
+ }
+ String entryName = null;
+ for (Map.Entry<String, Metadata> e : extracted.entrySet()) {
+ if (e.getValue() == thumbnail) {
+ entryName = e.getKey();
+ }
+ }
+ ZipEntry imageEntry = entryName == null ? null :
zip.getEntry(entryName);
+ if (imageEntry == null) {
+ throw new WebApplicationException(Response.Status.NO_CONTENT);
+ }
+ if (imageEntry.getSize() > MAX_THUMBNAIL_BYTES) {
+ throw new WebApplicationException("thumbnail larger than " +
MAX_THUMBNAIL_BYTES + " bytes",
+ Response.Status.REQUEST_ENTITY_TOO_LARGE);
+ }
+ byte[] image;
+ try (InputStream is = zip.getInputStream(imageEntry)) {
+ //the entry size is a claim; read one byte past the limit to
know
+ image = is.readNBytes((int) MAX_THUMBNAIL_BYTES + 1);
+ }
+ if (image.length > MAX_THUMBNAIL_BYTES) {
+ throw new WebApplicationException("thumbnail larger than " +
MAX_THUMBNAIL_BYTES + " bytes",
+ Response.Status.REQUEST_ENTITY_TOO_LARGE);
+ }
+ StringWriter metadataJson = new StringWriter();
+ JsonMetadata.toJson(thumbnail, metadataJson);
+ ObjectNode root = MAPPER.createObjectNode();
+ root.set("metadata", MAPPER.readTree(metadataJson.toString()));
+ root.put("image", Base64.getEncoder().encodeToString(image));
+ return
Response.ok(MAPPER.writeValueAsString(root)).type("application/json").build();
Review Comment:
Left as it is. `JsonMetadata` is the canonical serializer for Metadata and
only writes to a Writer, so nesting it under a field costs one parse; the
alternative is a raw-value write, which trades a small parse for a hole in the
type safety. Next to base64 of up to 32 MB of image bytes in the same response,
that parse is not what this endpoint spends its time on.
> /unpack/thumbnail: return the document thumbnail with its metadata
> ------------------------------------------------------------------
>
> Key: TIKA-4856
> URL: https://issues.apache.org/jira/browse/TIKA-4856
> Project: Tika
> Issue Type: New Feature
> Reporter: Dominik Schmidt
> Priority: Major
>
> With TIKA-4850 through TIKA-4855 every container format that carries a
> thumbnail emits it as a THUMBNAIL embedded document, the PDF parser renders
> pages as RENDERING documents, and the EMF/WMF renderer turns the vector
> thumbnails of Office documents into raster ones. Getting "the thumbnail of
> this file" out of that still takes format knowledge on the client: the
> THUMBNAIL of a Word or Excel file is an EMF/WMF whose usable form is the
> RENDERING underneath it, a PDF has no THUMBNAIL but a page RENDERING, the
> THUMBNAIL of a DOCX inside a ZIP is not the ZIP's, and with rendering enabled
> the picture of an embedded OLE object is a RENDERING too. Plus the request
> config that switches the renderers on.
> Proposal: POST /unpack/thumbnail next to /unpack and /unpack/all, multipart
> like them. It runs the usual forked parse in unpack mode with a fixed parse
> context (PDF page 1 rendered, EMF/WMF rendered) and picks, in this order: the
> raster THUMBNAIL at depth 1; the rendering of that thumbnail; the depth-1
> RENDERING of PDF page 1. The endpoint extracts what the document carries; it
> does not resize, convert or generate previews.
> The response is JSON: the /rmeta metadata object of the selected embedded
> document, and the image as base64. Thumbnails are small, so the encoding
> overhead does not matter, and the caller gets type, dimensions, origin
> (stored thumbnail or rendering, tk:rendering:rendered-by) and path in one
> round trip without unpacking a zip. 204 when the document has no thumbnail.
> {
> "metadata": {
> "Content-Type": "image/png",
> "Content-Length": "8459",
> "tiff:ImageWidth": "800",
> "tiff:ImageLength": "1131",
> "tk:embedded-resource-type": "RENDERING",
> "tk:embedded-resource-path": "/thumbnail.emf/thumbnail.png",
> "tk:embedded-depth": "2",
> "tk:rendering:rendered-by": "poi-metafile-renderer",
> "tk:resource-name": "thumbnail.png"
> },
> "image": "iVBORw0KGgoAAAANSUhEUgAA..."
> }
> To keep the selection rule short, the metafile renderer could give the
> rendering of a THUMBNAIL the THUMBNAIL type as well (its
> tk:rendering:rendered-by tells it apart), so a raster thumbnail is a
> THUMBNAIL regardless of whether the document stored it as PNG or as EMF.
> What do you think?
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
