Create a standalone pdfium build.

samples/pdfium_test.cc

Index: samples/pdfium_test.cc
diff --git a/samples/pdfium_test.cc b/samples/pdfium_test.cc
new file mode 100644
index 0000000000000000000000000000000000000000..6c14738a56783e2784b004e88159e8c847146840
--- /dev/null
+++ b/samples/pdfium_test.cc
@@ -0,0 +1,310 @@
+// Copyright (c) 2010 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#include <limits.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include <list>
+#include <string>
+#include <utility>
+
+#include "../fpdfsdk/include/fpdf_dataavail.h"

[Nico, 2014/05/23 21:14:33]

nit: Relative includes looks strange; maybe add `'include_dirs': ['<(DEPTH)'],`
to samples.gyp

[jam, 2014/05/23 23:52:13]

Done.

+#include "../fpdfsdk/include/fpdf_ext.h"
+#include "../fpdfsdk/include/fpdfformfill.h"
+#include "../fpdfsdk/include/fpdftext.h"
+#include "../fpdfsdk/include/fpdfview.h"
+#include "../v8/include/v8.h"
+
+#ifdef _WIN32
+  #define snprintf _snprintf

[Nico, 2014/05/23 21:14:33]

Not sure if it's important, but _snprintf is different from snprintf in how they
behave if the string is too long for the buffer: snprintf adds a trailing \0,
while _snprintf prefers to include one more character instead of adding the \0. 
(If this is important, I think you can use _snprintf_s in _TRUNCATE mode to
emulate snprintf on windows)

[jam, 2014/05/23 23:52:13]

i'm going to leave this since it's all moved code
(http://src.chromium.org/viewvc/chrome/trunk/src/pdf/pdfium/pdfium_test.cc?rev...)

+  /* in Windows, rb for open and read binary file */
+  #define FOPEN_READ "rb"
+#else
+  #define FOPEN_READ "r"
+#endif
+
+static void write_file(const char* pdf_name, int num,

[Nico, 2014/05/23 21:14:33]

nit: WritePpm?

[jam, 2014/05/23 23:52:13]

Done.

+                       const char* buffer, int stride, int width, int height) {
+  if (stride < 0 || width < 0 || height < 0)
+    return;
+  if (height > 0 && width > INT_MAX / height)
+    return;
+  int out_len = width * height;
+  if (out_len > INT_MAX / 3)
+    return;
+  out_len *= 3;
+
+  char filename[256];
+  snprintf(filename, sizeof(filename), "%s.%d.ppm", pdf_name, num);
+  FILE* fp = fopen(filename, "w");
+  if (!fp)
+    return;
+  fprintf(fp, "P6\n# PDF test render\n%d %d\n255\n", width, height);
+  // Source data is B, G, R, unused.
+  // Dest data is R, G, B.
+  char* result = new char[out_len];
+  if (result) {
+    int h, w;

[Nico, 2014/05/23 21:14:33]

declare these in the for loop, you only use them in the respective loop bodies
(`for (int h = 0; …`)

[jam, 2014/05/23 23:52:13]

Done.

+    for (h = 0; h < height; ++h) {
+      const char* src_line = buffer + (stride * h);
+      char* dest_line = result + (width * h * 3);
+      for (w = 0; w < width; ++w) {
+        // R
+        dest_line[w * 3] = src_line[(w * 4) + 2];
+        // G
+        dest_line[(w * 3) + 1] = src_line[(w * 4) + 1];
+        // B
+        dest_line[(w * 3) + 2] = src_line[w * 4];
+      }
+    }
+    fwrite(result, out_len, 1, fp);
+    delete [] result;
+  }
+  fclose(fp);
+}

[Nico, 2014/05/23 21:14:33]

(fwiw, writing an inefficiently compressed png isn't much harder:
https://github.com/nico/hack/blob/master/wpng/wpng.c and if you're going for
easy-to-write, consider tga instead of ppm:
https://github.com/nico/hack/blob/master/wpng/wtga.c)

[jam, 2014/05/23 23:52:13]

ditto

+
+int Form_Alert(IPDF_JSPLATFORM*, FPDF_WIDESTRING, FPDF_WIDESTRING, int, int) {

[Nico, 2014/05/23 21:14:33]

nit: no underscores in function names (?) (if so, also below)

[jam, 2014/05/23 23:52:13]

ditto

+  printf("Form_Alert called.\n");

[Nico, 2014/05/23 21:14:33]

should this print to stderr?

[jam, 2014/05/23 23:52:13]

ditto

[jam, 2014/05/23 23:52:13]

ditto (no idea about these questions, this tool is used by fuzzing)

+  return 0;
+}
+
+void Unsupported_Handler(UNSUPPORT_INFO*, int type) {
+  std::string feature = "Unknown";

[Nico, 2014/05/23 21:14:33]

(nit: i'd make this a const char*, then there's fewer copying and the printf
line is slightly shorter)

[jam, 2014/05/23 23:52:13]

ditto

+  switch (type) {
+    case FPDF_UNSP_DOC_XFAFORM:
+      feature = "XFA";
+      break;
+    case FPDF_UNSP_DOC_PORTABLECOLLECTION:
+      feature = "Portfolios_Packages";
+      break;
+    case FPDF_UNSP_DOC_ATTACHMENT:
+    case FPDF_UNSP_ANNOT_ATTACHMENT:
+      feature = "Attachment";
+      break;
+    case FPDF_UNSP_DOC_SECURITY:
+      feature = "Rights_Management";
+      break;
+    case FPDF_UNSP_DOC_SHAREDREVIEW:
+      feature = "Shared_Review";
+      break;
+    case FPDF_UNSP_DOC_SHAREDFORM_ACROBAT:
+    case FPDF_UNSP_DOC_SHAREDFORM_FILESYSTEM:
+    case FPDF_UNSP_DOC_SHAREDFORM_EMAIL:
+      feature = "Shared_Form";
+      break;
+    case FPDF_UNSP_ANNOT_3DANNOT:
+      feature = "3D";
+      break;
+    case FPDF_UNSP_ANNOT_MOVIE:
+      feature = "Movie";
+      break;
+    case FPDF_UNSP_ANNOT_SOUND:
+      feature = "Sound";
+      break;
+    case FPDF_UNSP_ANNOT_SCREEN_MEDIA:
+    case FPDF_UNSP_ANNOT_SCREEN_RICHMEDIA:
+      feature = "Screen";
+      break;
+    case FPDF_UNSP_ANNOT_SIG:
+      feature = "Digital_Signature";
+      break;
+  }
+  printf("Unsupported feature: %s.\n", feature.c_str());

[Nico, 2014/05/23 21:14:33]

stderr?

+}
+
+bool ParseCommandLine(int argc, const char* argv[], bool* write_images,
+                      std::list<const char*>* files) {

[Nico, 2014/05/23 21:14:33]

nit: use a vector instead of a list (then the main loop can just be over an int
too)

+  *write_images = false;
+  files->clear();
+
+  int cur_arg = 1;
+  if (cur_arg < argc &&
+      strcmp(argv[cur_arg], "--write_images") == 0) {
+    *write_images = true;
+    cur_arg++;
+  }
+
+  if (cur_arg >= argc)
+    return false;
+
+  for (int i = cur_arg; i < argc; i++)
+    files->push_back(argv[i]);
+
+  return true;
+}
+
+class TestLoader {
+ public:
+  TestLoader(const char* pBuf, size_t len);
+
+  const char* m_pBuf;
+  size_t m_Len;
+};
+
+TestLoader::TestLoader(const char* pBuf, size_t len)
+    : m_pBuf(pBuf), m_Len(len) {
+}
+
+int Get_Block(void* param, unsigned long pos, unsigned char* pBuf,
+              unsigned long size) {
+  TestLoader* pLoader = (TestLoader*) param;
+  if (pos + size < pos || pos + size > pLoader->m_Len) return 0;
+  memcpy(pBuf, pLoader->m_pBuf + pos, size);
+  return 1;
+}
+
+bool Is_Data_Avail(FX_FILEAVAIL* pThis, size_t offset, size_t size) {
+  return true;
+}
+
+void Add_Segment(FX_DOWNLOADHINTS* pThis, size_t offset, size_t size) {
+}
+
+void RenderPdf(const char* name, const char* pBuf, size_t len,
+               bool write_images) {
+  printf("Rendering PDF file %s.\n", name);
+
+  IPDF_JSPLATFORM platform_callbacks;
+  memset(&platform_callbacks, '\0', sizeof(platform_callbacks));

[Nico, 2014/05/23 21:14:33]

nit: 0 is more common than \0 in memset calls (also below)

+  platform_callbacks.version = 1;
+  platform_callbacks.app_alert = Form_Alert;
+
+  FPDF_FORMFILLINFO form_callbacks;
+  memset(&form_callbacks, '\0', sizeof(form_callbacks));
+  form_callbacks.version = 1;
+  form_callbacks.m_pJsPlatform = &platform_callbacks;
+
+  TestLoader loader(pBuf, len);
+
+  FPDF_FILEACCESS file_access;
+  memset(&file_access, '\0', sizeof(file_access));
+  file_access.m_FileLen = len;
+  file_access.m_GetBlock = Get_Block;
+  file_access.m_Param = &loader;
+
+  FX_FILEAVAIL file_avail;
+  memset(&file_avail, '\0', sizeof(file_avail));
+  file_avail.version = 1;
+  file_avail.IsDataAvail = Is_Data_Avail;
+
+  FX_DOWNLOADHINTS hints;
+  memset(&hints, '\0', sizeof(hints));
+  hints.version = 1;
+  hints.AddSegment = Add_Segment;
+
+  FPDF_DOCUMENT doc;
+  FPDF_AVAIL pdf_avail = FPDFAvail_Create(&file_avail, &file_access);
+
+  (void) FPDFAvail_IsDocAvail(pdf_avail, &hints);

[Nico, 2014/05/23 21:14:33]

Maybe add a comment why this is being called; it's not clear to me

+
+  if (!FPDFAvail_IsLinearized(pdf_avail)) {
+    printf("Non-linearized path...\n");
+    doc = FPDF_LoadCustomDocument(&file_access, NULL);
+  } else {
+    printf("Linearized path...\n");
+    doc = FPDFAvail_GetDocument(pdf_avail, NULL);
+  }
+
+  (void) FPDF_GetDocPermissions(doc);
+  (void) FPDFAvail_IsFormAvail(pdf_avail, &hints);
+
+  FPDF_FORMHANDLE form = FPDFDOC_InitFormFillEnviroument(doc, &form_callbacks);
+  FPDF_SetFormFieldHighlightColor(form, 0, 0xFFE4DD);
+  FPDF_SetFormFieldHighlightAlpha(form, 100);

[Nico, 2014/05/23 21:14:33]

nit: maybe use a number that isn't 100 here. The argument goes from 0-255, and
100 suggests it might be a percentage thing. Maybe `, 142);  // arbitrary)

or write 100 as 0x64, then it doesn't look like a percentage either

+
+  int first_page = FPDFAvail_GetFirstPageNum(doc);
+  (void) FPDFAvail_IsPageAvail(pdf_avail, first_page, &hints);
+
+  int page_count = FPDF_GetPageCount(doc);
+  for (int i = 0; i < page_count; ++i) {
+    (void) FPDFAvail_IsPageAvail(pdf_avail, i, &hints);
+  }
+
+  FORM_DoDocumentJSAction(form);
+  FORM_DoDocumentOpenAction(form);
+
+  for (int i = 0; i < page_count; ++i) {
+    FPDF_PAGE page = FPDF_LoadPage(doc, i);
+    FPDF_TEXTPAGE text_page = FPDFText_LoadPage(page);
+    FORM_OnAfterLoadPage(page, form);
+    FORM_DoPageAAction(page, form, FPDFPAGE_AACTION_OPEN);
+
+    int width = static_cast<int>(FPDF_GetPageWidth(page));
+    int height = static_cast<int>(FPDF_GetPageHeight(page));
+    FPDF_BITMAP bitmap = FPDFBitmap_Create(width, height, 0);
+    FPDFBitmap_FillRect(bitmap, 0, 0, width, height, 255, 255, 255, 255);
+
+    FPDF_RenderPageBitmap(bitmap, page, 0, 0, width, height, 0, 0);
+    FPDF_FFLDraw(form, bitmap, page, 0, 0, width, height, 0, 0);
+    if (write_images) {
+      const char* buffer = reinterpret_cast<const char*>(
+          FPDFBitmap_GetBuffer(bitmap));
+      int stride = FPDFBitmap_GetStride(bitmap);
+      write_file(name, i, buffer, stride, width, height);
+    }
+
+    FPDFBitmap_Destroy(bitmap);
+
+    FORM_DoPageAAction(page, form, FPDFPAGE_AACTION_CLOSE);
+    FORM_OnBeforeClosePage(page, form);
+    FPDFText_ClosePage(text_page);
+    FPDF_ClosePage(page);
+  }
+
+  FORM_DoDocumentAAction(form, FPDFDOC_AACTION_WC);
+  FPDFDOC_ExitFormFillEnviroument(form);
+  FPDF_CloseDocument(doc);
+  FPDFAvail_Destroy(pdf_avail);
+
+  printf("Loaded, parsed and rendered %d pages.\n", page_count);
+}
+
+int main(int argc, const char* argv[]) {
+  v8::V8::InitializeICU();
+  bool write_images = false;
+  std::list<const char*> files;
+  if (!ParseCommandLine(argc, argv, &write_images, &files)) {
+    printf("Usage is: test [--write_images] /path/to/pdf\n");
+    printf("--write_images - to write page images <pdf-name>.<page-number>.ppm\n");
+    return 1;
+  }
+
+  FPDF_InitLibrary(NULL);
+
+  UNSUPPORT_INFO unsuppored_info;
+  memset(&unsuppored_info, '\0', sizeof(unsuppored_info));
+  unsuppored_info.version = 1;
+  unsuppored_info.FSDK_UnSupport_Handler = Unsupported_Handler;
+
+  FSDK_SetUnSpObjProcessHandler(&unsuppored_info);
+
+  while (!files.empty()) {
+    const char* filename = files.front();
+    files.pop_front();
+    FILE* file = fopen(filename, FOPEN_READ);
+    if (!file) {
+      fprintf(stderr, "Failed to open: %s\n", filename);
+      continue;
+    }
+    (void) fseek(file, 0, SEEK_END);
+    size_t len = ftell(file);
+    (void) fseek(file, 0, SEEK_SET);
+    char* pBuf = (char*) malloc(len);
+    size_t ret = fread(pBuf, 1, len, file);
+    (void) fclose(file);
+    if (ret != len) {
+      fprintf(stderr, "Failed to read: %s\n", filename);
+    } else {
+      RenderPdf(filename, pBuf, len, write_images);
+    }
+    free(pBuf);
+  }
+
+  FPDF_DestroyLibrary();
+
+  return 0;
+}
+