From 7ab4e53b99bee4ec56c54f8bef4072f6216171ed Mon Sep 17 00:00:00 2001
From: Nico Weber <thakis@chromium.org>
Date: Fri, 9 Feb 2024 08:08:00 -0500
Subject: [PATCH] LibPDF/CFF: Add code for fdselect parsing

This is one of the two top dict entries we need for CID-keyed fonts.
We don't send any CID-keyed font data into the CFF parser yet,
so no behavior change.
---
 Userland/Libraries/LibPDF/Fonts/CFF.cpp | 57 ++++++++++++++++++++++++-
 Userland/Libraries/LibPDF/Fonts/CFF.h   |  1 +
 2 files changed, 57 insertions(+), 1 deletion(-)

diff --git a/Userland/Libraries/LibPDF/Fonts/CFF.cpp b/Userland/Libraries/LibPDF/Fonts/CFF.cpp
index 1234cb7d2e..2d9575bc1a 100644
--- a/Userland/Libraries/LibPDF/Fonts/CFF.cpp
+++ b/Userland/Libraries/LibPDF/Fonts/CFF.cpp
@@ -128,6 +128,7 @@ PDFErrorOr<NonnullRefPtr<CFF>> CFF::create(ReadonlyBytes const& cff_bytes, RefPt
     Vector<ByteBuffer> local_subroutines;
     float defaultWidthX = 0;
     float nominalWidthX = 0;
+    int fdselect_offset = 0;
     TRY(parse_index(reader, [&](ReadonlyBytes const& element_data) {
         Reader element_reader { element_data };
         return parse_dict<TopDictOperator>(element_reader, [&](TopDictOperator op, Vector<DictOperand> const& operands) -> PDFErrorOr<void> {
@@ -229,13 +230,16 @@ PDFErrorOr<NonnullRefPtr<CFF>> CFF::create(ReadonlyBytes const& cff_bytes, RefPt
                 }));
                 break;
             }
+            case TopDictOperator::FDSelect:
+                if (!operands.is_empty())
+                    fdselect_offset = operands[0].get<int>();
+                break;
             case TopDictOperator::CIDFontVersion:
             case TopDictOperator::CIDFontRevision:
             case TopDictOperator::CIDFontType:
             case TopDictOperator::CIDCount:
             case TopDictOperator::UIDBase:
             case TopDictOperator::FDArray:
-            case TopDictOperator::FDSelect:
             case TopDictOperator::FontName:
                 // Keys for CID-keyed fonts that we don't need, at least at the moment.
                 break;
@@ -306,6 +310,12 @@ PDFErrorOr<NonnullRefPtr<CFF>> CFF::create(ReadonlyBytes const& cff_bytes, RefPt
     }
     }
 
+    // CFF spec, "19 FDSelect"
+    if (fdselect_offset != 0) {
+        auto fdselect = TRY(parse_fdselect(Reader { cff_bytes.slice(fdselect_offset) }, glyphs.size()));
+        dbgln_if(CFF_DEBUG, "CFF has {} FDSelect entries", fdselect.size());
+    }
+
     // Adjust glyphs' widths as they are deltas from nominalWidthX
     for (auto& glyph : glyphs) {
         if (!glyph.has_width())
@@ -807,6 +817,51 @@ PDFErrorOr<Vector<CFF::SID>> CFF::parse_charset(Reader&& reader, size_t glyph_co
     return names;
 }
 
+PDFErrorOr<Vector<u8>> CFF::parse_fdselect(Reader&& reader, size_t glyph_count)
+{
+    Vector<u8> fd_selector_array; // Maps GIDs to their FD index.
+
+    // CFF spec, "19 FDSelect"
+    auto format = TRY(reader.try_read<Card8>());
+
+    if (format == 0) {
+        // CFF spec, "Table 27 Format 0"
+        // "(This format is identical to charset format 0 except that the notdef glyph is included in this case.)"
+        dbgln_if(CFF_DEBUG, "CFF fdselect format 0");
+        fd_selector_array.ensure_capacity(glyph_count);
+        for (size_t i = 0; i < glyph_count; i++)
+            fd_selector_array.append(TRY(reader.try_read<Card8>()));
+    } else if (format == 3) {
+        // CFF spec, "Table 28 Format 3"
+        dbgln_if(CFF_DEBUG, "CFF fdselect format 3");
+
+        // The spec presents this as n "Card16 first; Card8 fd;" struct entries followed by a Char16 sentinel value.
+        // But the code is shorter if we treat it as a Char16 start value followed by n "Card8 fd; Card16 end;" struct entries.
+        Card16 n_ranges = TRY(reader.try_read<BigEndian<Card16>>());
+        Card16 begin = TRY(reader.try_read<BigEndian<Card16>>());
+
+        // "The first range must have a 'first' GID of 0."
+        if (begin != 0)
+            return error("CFF fdselect format 3 first range must have a 'first' GID of 0");
+
+        for (Card16 i = 0; i < n_ranges; i++) {
+            auto fd = TRY(reader.try_read<Card8>());
+            auto end = TRY(reader.try_read<BigEndian<Card16>>());
+            for (Card16 j = begin; j < end; j++)
+                fd_selector_array.append(fd);
+            begin = end;
+        }
+
+        // "The sentinel GID is set equal to the number of glyphs in the font."
+        if (begin != glyph_count)
+            return error("CFF fdselect format 3 last range must end at the number of glyphs in the font");
+    } else {
+        dbgln("CFF: Unknown fdselect format {}", format);
+    }
+
+    return fd_selector_array;
+}
+
 PDFErrorOr<Vector<CFF::Glyph>> CFF::parse_charstrings(Reader&& reader, Vector<ByteBuffer> const& local_subroutines, Vector<ByteBuffer> const& global_subroutines)
 {
     // CFF spec, "14 CharStrings INDEX"
diff --git a/Userland/Libraries/LibPDF/Fonts/CFF.h b/Userland/Libraries/LibPDF/Fonts/CFF.h
index feac924d74..ad7bf9d9d6 100644
--- a/Userland/Libraries/LibPDF/Fonts/CFF.h
+++ b/Userland/Libraries/LibPDF/Fonts/CFF.h
@@ -124,6 +124,7 @@ public:
 
     static DeprecatedFlyString resolve_sid(SID, Vector<StringView> const&);
     static PDFErrorOr<Vector<SID>> parse_charset(Reader&&, size_t);
+    static PDFErrorOr<Vector<u8>> parse_fdselect(Reader&&, size_t);
     static PDFErrorOr<Vector<u8>> parse_encoding(Reader&&, HashMap<Card8, SID>& supplemental);
 };