Add metric collection for form type in PDFs

BUG=chromium:775519,chromium:768986 Change-Id: I485e4ec14fa2097f8d0ef7ff782ac2d086d78411 Reviewed-on: https://chromium-review.googlesource.com/723581 Commit-Queue: Ryan Harrison <rharrison@chromium.org> Reviewed-by: Lei Zhang <thestig@chromium.org> Reviewed-by: Ilya Sherman <isherman@chromium.org> Cr-Commit-Position: refs/heads/master@{#510827}
Denger-Network · Oct 23, 2017 · 4de8f1e · 4de8f1e
1 parent 8cddb0d
commit 4de8f1e
Show file tree

Hide file tree

Showing 5 changed files with 45 additions and 0 deletions.
diff --git a/pdf/out_of_process_instance.cc b/pdf/out_of_process_instance.cc
@@ -1506,6 +1506,9 @@ void OutOfProcessInstance::DocumentLoadComplete(
   HistogramEnumeration("PDF.IsLinearized",
                        document_features.is_linearized ? 1 : 0, 2);
   HistogramEnumeration("PDF.IsTagged", document_features.is_tagged ? 1 : 0, 2);
+  HistogramEnumeration("PDF.FormType",
+                       static_cast<int32_t>(document_features.form_type),
+                       static_cast<int32_t>(PDFEngine::FormType::kCount));
 }
 
 void OutOfProcessInstance::RotateClockwise() {

diff --git a/pdf/pdf_engine.h b/pdf/pdf_engine.h
@@ -58,6 +58,16 @@ class PDFEngine {
     PERMISSION_PRINT_HIGH_QUALITY,
   };
 
+  // Values other then |kCount| are persisted to logs as part of metric
+  // collection, so should not be changed.
+  enum class FormType {
+    kNone = 0,
+    kAcroForm = 1,
+    kXFAFull = 2,
+    kXFAForeground = 3,
+    kCount = 4,
+  };
+
   struct DocumentFeatures {
     // Number of pages in document.
     size_t page_count = 0;
@@ -69,6 +79,8 @@ class PDFEngine {
     bool is_linearized = false;
     // Whether the PDF is Tagged (see 10.7 "Tagged PDF" in PDF Reference 1.7).
     bool is_tagged = false;
+    // What type of form the document contains.
+    FormType form_type = FormType::kNone;
   };
 
   // The interface that's provided to the rendering engine.

diff --git a/pdf/pdfium/pdfium_engine.cc b/pdf/pdfium/pdfium_engine.cc
@@ -74,6 +74,20 @@ using printing::kPixelsPerInch;
 
 namespace chrome_pdf {
 
+static_assert(static_cast<int>(PDFEngine::FormType::kNone) == FORMTYPE_NONE,
+              "None form types must match");
+static_assert(static_cast<int>(PDFEngine::FormType::kAcroForm) ==
+                  FORMTYPE_ACRO_FORM,
+              "AcroForm form types must match");
+static_assert(static_cast<int>(PDFEngine::FormType::kXFAFull) ==
+                  FORMTYPE_XFA_FULL,
+              "XFA full form types must match");
+static_assert(static_cast<int>(PDFEngine::FormType::kXFAForeground) ==
+                  FORMTYPE_XFA_FOREGROUND,
+              "XFA foreground form types must match");
+static_assert(static_cast<int>(PDFEngine::FormType::kCount) == FORMTYPE_COUNT,
+              "Form type counts must match");
+
 namespace {
 
 const int32_t kPageShadowTop = 3;
@@ -1282,6 +1296,7 @@ void PDFiumEngine::FinishLoadingDocument() {
     document_features.is_linearized =
         (FPDFAvail_IsLinearized(fpdf_availability_) == PDF_LINEARIZED);
     document_features.is_tagged = FPDFCatalog_IsTagged(doc_);
+    document_features.form_type = static_cast<FormType>(FPDF_GetFormType(doc_));
     client_->DocumentLoadComplete(document_features);
   }
 }

diff --git a/tools/metrics/histograms/enums.xml b/tools/metrics/histograms/enums.xml
@@ -31229,6 +31229,13 @@ from previous Chrome versions.
   <int value="2" label="Has Bookmarks"/>
 </enum>
 
+<enum name="PDFFormTypes">
+  <int value="0" label="None"/>
+  <int value="1" label="AcroForm"/>
+  <int value="2" label="Full XFA"/>
+  <int value="3" label="Foreground XFA (XFAF)"/>
+</enum>
+
 <enum name="PeerConnectionCounters">
   <int value="0" label="PeerConnection enabled with IPv4."/>
   <int value="1" label="PeerConnection enabled with Ipv6."/>

diff --git a/tools/metrics/histograms/histograms.xml b/tools/metrics/histograms/histograms.xml
@@ -58368,6 +58368,14 @@ http://cs/file:chrome/histograms.xml - but prefer this file for new entries.
   </summary>
 </histogram>
 
+<histogram name="PDF.FormType" enum="PDFFormTypes">
+  <owner>rharrison@chromium.org</owner>
+  <summary>
+    Tracks what types of forms are present in PDF document, logged when the
+    document finishes loading.
+  </summary>
+</histogram>
+
 <histogram name="PDF.HasAttachment" enum="Boolean">
   <owner>hnakashima@chromium.org</owner>
   <summary>