Name the states for fpdf_parser_parser.

core/src/fpdfapi/fpdf_parser/fpdf_parser_parser.cpp

Index: core/src/fpdfapi/fpdf_parser/fpdf_parser_parser.cpp
diff --git a/core/src/fpdfapi/fpdf_parser/fpdf_parser_parser.cpp b/core/src/fpdfapi/fpdf_parser/fpdf_parser_parser.cpp
index 9d646b2943e140c0f6764b5b61c04e65c861239a..189d070ff83e9aecfe9e9b35e6471e5c44a5d155 100644
--- a/core/src/fpdfapi/fpdf_parser/fpdf_parser_parser.cpp
+++ b/core/src/fpdfapi/fpdf_parser/fpdf_parser_parser.cpp
@@ -36,6 +36,27 @@ struct SearchTagRecord {
   FX_DWORD m_Offset;
 };
 
+enum class ParserState {

[Tom Sepez, 2016/02/18 19:27:55]

Can we make ParserState an enum nested privately in CPDF_Parser?

[dsinclair, 2016/02/18 19:47:04]

Done.

Is there a benefit over having it in the anonymous namespace?

[Tom Sepez, 2016/02/18 19:49:16]

Yes, it tells the reader that this state is associated with the parser class and
only the parser class.

[dsinclair, 2016/02/18 19:51:03]

Ah, right. Too spoiled by one class per file. Wasn't thinking about the
many-classes per file we have.

+  kDefault,
+  kComment,

[Tom Sepez, 2016/02/18 19:27:55]

Can we keep these in the same order as before so that the values of the states
don't get permuted?  e.g. kWhiteSpace as 1?

I'm worried that in the depths of the logic, they may do some bizzaro thing like
if state > N or some such.

[dsinclair, 2016/02/18 19:47:04]

I checked through the code, all uses of status are to assign the new state or in
the switch(status). 

Putting in the existing order will force us to put in a dummy 12 state, which we
don't need.

[Tom Sepez, 2016/02/18 19:49:16]

Acknowledged.

+  kWhitespace,
+

[Tom Sepez, 2016/02/18 19:27:55]

nit: not sure we need blank lines here.

[dsinclair, 2016/02/18 19:47:04]

Done. Was an attempt at grouping and readability.

+  kString,
+  kHexString,
+  kEscapedString,
+
+  kXref,
+  kObjNum,
+  kPostObjNum,
+  kGenNum,
+  kPostGenNum,
+
+  kTrailer,
+
+  kBeginObj,
+  kEndObj
+};
+
 int32_t GetHeaderOffset(IFX_FileRead* pFile) {
   const FX_DWORD tag = FXDWORD_FROM_LSBFIRST(0x46445025);
   const size_t kBufSize = 4;
@@ -582,7 +603,9 @@ FX_BOOL CPDF_Parser::RebuildCrossRef() {
     m_pTrailer->Release();
     m_pTrailer = NULL;
   }
-  int32_t status = 0;
+
+  enum ParserState status = ParserState::kDefault;

[Tom Sepez, 2016/02/18 19:27:55]

nit: |enum| needed ? Also, can we rename this |state| since "status" means
something slightly different (to me) -- its a "state machine", not a "status
machine". :).

[dsinclair, 2016/02/18 19:47:04]

Done.

+
   int32_t inside_index = 0;
   FX_DWORD objnum = 0;
   FX_DWORD gennum = 0;
@@ -605,97 +628,105 @@ FX_BOOL CPDF_Parser::RebuildCrossRef() {
     for (FX_DWORD i = 0; i < size; i++) {
       uint8_t byte = buffer[i];
       switch (status) {
-        case 0:
+        case ParserState::kDefault:
           if (PDFCharIsWhitespace(byte))
-            status = 1;
+            status = ParserState::kWhitespace;
 
           if (std::isdigit(byte)) {
             --i;
-            status = 1;
+            status = ParserState::kWhitespace;
           }
 
           if (byte == '%') {
             inside_index = 0;
-            status = 9;
+            status = ParserState::kComment;
           }
 
           if (byte == '(') {
-            status = 10;
+            status = ParserState::kString;
             depth = 1;
           }
 
           if (byte == '<') {
             inside_index = 1;
-            status = 11;
+            status = ParserState::kHexString;
           }
 
           if (byte == '\\')
-            status = 13;
+            status = ParserState::kEscapedString;
 
           if (byte == 't') {
-            status = 7;
+            status = ParserState::kTrailer;
             inside_index = 1;
           }
           break;
-        case 1:
+
+        case ParserState::kWhitespace:
           if (PDFCharIsWhitespace(byte)) {
             break;
           } else if (std::isdigit(byte)) {
             start_pos = pos + i;
-            status = 2;
+            status = ParserState::kObjNum;
             objnum = FXSYS_toDecimalDigit(byte);
+
           } else if (byte == 't') {
-            status = 7;
+            status = ParserState::kTrailer;
             inside_index = 1;
+
           } else if (byte == 'x') {
-            status = 8;
+            status = ParserState::kXref;
             inside_index = 1;
+
           } else {
             --i;
-            status = 0;
+            status = ParserState::kDefault;
           }
           break;
-        case 2:
+
+        case ParserState::kObjNum:
           if (std::isdigit(byte)) {
             objnum = objnum * 10 + FXSYS_toDecimalDigit(byte);
             break;
           } else if (PDFCharIsWhitespace(byte)) {
-            status = 3;
+            status = ParserState::kPostObjNum;
           } else {
             --i;
-            status = 14;
+            status = ParserState::kEndObj;
             inside_index = 0;
           }
           break;
-        case 3:
+
+        case ParserState::kPostObjNum:
           if (std::isdigit(byte)) {
             start_pos1 = pos + i;
-            status = 4;
+            status = ParserState::kGenNum;
             gennum = FXSYS_toDecimalDigit(byte);
           } else if (PDFCharIsWhitespace(byte)) {
             break;
           } else if (byte == 't') {
-            status = 7;
+            status = ParserState::kTrailer;
             inside_index = 1;
           } else {
             --i;
-            status = 0;
+            status = ParserState::kDefault;
           }
           break;
-        case 4:
+
+        case ParserState::kGenNum:
           if (std::isdigit(byte)) {
             gennum = gennum * 10 + FXSYS_toDecimalDigit(byte);
             break;
           } else if (PDFCharIsWhitespace(byte)) {
-            status = 5;
+            status = ParserState::kPostGenNum;
           } else {
             --i;
-            status = 0;
+            status = ParserState::kDefault;
           }
           break;
-        case 5:
+
+        case ParserState::kPostGenNum:
           if (byte == 'o') {
-            status = 6;
+            status = ParserState::kBeginObj;
             inside_index = 1;
           } else if (PDFCharIsWhitespace(byte)) {
             break;
@@ -704,21 +735,22 @@ FX_BOOL CPDF_Parser::RebuildCrossRef() {
             gennum = FXSYS_toDecimalDigit(byte);
             start_pos = start_pos1;
             start_pos1 = pos + i;
-            status = 4;
+            status = ParserState::kGenNum;
           } else if (byte == 't') {
-            status = 7;
+            status = ParserState::kTrailer;
             inside_index = 1;
           } else {
             --i;
-            status = 0;
+            status = ParserState::kDefault;
           }
           break;
-        case 6:
+
+        case ParserState::kBeginObj:
           switch (inside_index) {
             case 1:
               if (byte != 'b') {
                 --i;
-                status = 0;
+                status = ParserState::kDefault;
               } else {
                 inside_index++;
               }
@@ -726,7 +758,7 @@ FX_BOOL CPDF_Parser::RebuildCrossRef() {
             case 2:
               if (byte != 'j') {
                 --i;
-                status = 0;
+                status = ParserState::kDefault;
               } else {
                 inside_index++;
               }
@@ -734,7 +766,7 @@ FX_BOOL CPDF_Parser::RebuildCrossRef() {
             case 3:
               if (PDFCharIsWhitespace(byte) || PDFCharIsDelimiter(byte)) {
                 if (objnum > 0x1000000) {
-                  status = 0;
+                  status = ParserState::kDefault;
                   break;
                 }
                 FX_FILESIZE obj_pos = start_pos - m_Syntax.m_HeaderOffset;
@@ -793,11 +825,12 @@ FX_BOOL CPDF_Parser::RebuildCrossRef() {
                 }
               }
               --i;
-              status = 0;
+              status = ParserState::kDefault;
               break;
           }
           break;
-        case 7:
+
+        case ParserState::kTrailer:
           if (inside_index == 7) {
             if (PDFCharIsWhitespace(byte) || PDFCharIsDelimiter(byte)) {
               last_trailer = pos + i - 7;
@@ -860,31 +893,34 @@ FX_BOOL CPDF_Parser::RebuildCrossRef() {
               }
             }
             --i;
-            status = 0;
+            status = ParserState::kDefault;
           } else if (byte == "trailer"[inside_index]) {
             inside_index++;
           } else {
             --i;
-            status = 0;
+            status = ParserState::kDefault;
           }
           break;
-        case 8:
+
+        case ParserState::kXref:
           if (inside_index == 4) {
             last_xref = pos + i - 4;
-            status = 1;
+            status = ParserState::kWhitespace;
           } else if (byte == "xref"[inside_index]) {
             inside_index++;
           } else {
             --i;
-            status = 0;
+            status = ParserState::kDefault;
           }
           break;
-        case 9:
+
+        case ParserState::kComment:
           if (byte == '\r' || byte == '\n') {
-            status = 0;
+            status = ParserState::kDefault;
           }
           break;
-        case 10:
+
+        case ParserState::kString:
           if (byte == ')') {
             if (depth > 0) {
               depth--;
@@ -893,29 +929,32 @@ FX_BOOL CPDF_Parser::RebuildCrossRef() {
             depth++;
           }
           if (!depth) {
-            status = 0;
+            status = ParserState::kDefault;
           }
           break;
-        case 11:
+
+        case ParserState::kHexString:
           if (byte == '>' || (byte == '<' && inside_index == 1))
-            status = 0;
+            status = ParserState::kDefault;
           inside_index = 0;
           break;
-        case 13:

[Tom Sepez, 2016/02/18 19:27:55]

Heh.  No 12.

[dsinclair, 2016/02/18 19:47:04]

Acknowledged.

+
+        case ParserState::kEscapedString:
           if (PDFCharIsDelimiter(byte) || PDFCharIsWhitespace(byte)) {
             --i;
-            status = 0;
+            status = ParserState::kDefault;
           }
           break;
-        case 14:
+
+        case ParserState::kEndObj:
           if (PDFCharIsWhitespace(byte)) {
-            status = 0;
+            status = ParserState::kDefault;
           } else if (byte == '%' || byte == '(' || byte == '<' ||
                      byte == '\\') {
-            status = 0;
+            status = ParserState::kDefault;
             --i;
           } else if (inside_index == 6) {
-            status = 0;
+            status = ParserState::kDefault;
             --i;
           } else if (byte == "endobj"[inside_index]) {
             inside_index++;