First commit

1 files changed, 142 insertions, 0 deletions

diff --git a/pkg/gitdiff/parser.go b/pkg/gitdiff/parser.go
new file mode 100644
index 0000000..e8f8430
--- /dev/null
+++ b/pkg/gitdiff/parser.go
@@ -0,0 +1,142 @@
+// Package gitdiff parses and applies patches generated by Git. It supports
+// line-oriented text patches, binary patches, and can also parse standard
+// unified diffs generated by other tools.
+package gitdiff
+
+import (
+	"bufio"
+	"fmt"
+	"io"
+)
+
+// Parse parses a patch with changes to one or more files. Any content before
+// the first file is returned as the second value. If an error occurs while
+// parsing, it returns all files parsed before the error.
+//
+// Parse expects to receive a single patch. If the input may contain multiple
+// patches (for example, if it is an mbox file), callers should split it into
+// individual patches and call Parse on each one.
+func Parse(r io.Reader) ([]*File, string, error) {
+	p := newParser(r)
+
+	if err := p.Next(); err != nil {
+		if err == io.EOF {
+			return nil, "", nil
+		}
+		return nil, "", err
+	}
+
+	var preamble string
+	var files []*File
+	for {
+		file, pre, err := p.ParseNextFileHeader()
+		if err != nil {
+			return files, preamble, err
+		}
+		if len(files) == 0 {
+			preamble = pre
+		}
+		if file == nil {
+			break
+		}
+
+		for _, fn := range []func(*File) (int, error){
+			p.ParseTextFragments,
+			p.ParseBinaryFragments,
+		} {
+			n, err := fn(file)
+			if err != nil {
+				return files, preamble, err
+			}
+			if n > 0 {
+				break
+			}
+		}
+
+		files = append(files, file)
+	}
+
+	return files, preamble, nil
+}
+
+// TODO(bkeyes): consider exporting the parser type with configuration
+// this would enable OID validation, p-value guessing, and prefix stripping
+// by allowing users to set or override defaults
+
+// parser invariants:
+// - methods that parse objects:
+//     - start with the parser on the first line of the first object
+//     - if returning nil, do not advance
+//     - if returning an error, do not advance past the object
+//     - if returning an object, advance to the first line after the object
+// - any exported parsing methods must initialize the parser by calling Next()
+
+type stringReader interface {
+	ReadString(delim byte) (string, error)
+}
+
+type parser struct {
+	r stringReader
+
+	eof    bool
+	lineno int64
+	lines  [3]string
+}
+
+func newParser(r io.Reader) *parser {
+	if r, ok := r.(stringReader); ok {
+		return &parser{r: r}
+	}
+	return &parser{r: bufio.NewReader(r)}
+}
+
+// Next advances the parser by one line. It returns any error encountered while
+// reading the line, including io.EOF when the end of stream is reached.
+func (p *parser) Next() error {
+	if p.eof {
+		return io.EOF
+	}
+
+	if p.lineno == 0 {
+		// on first call to next, need to shift in all lines
+		for i := 0; i < len(p.lines)-1; i++ {
+			if err := p.shiftLines(); err != nil && err != io.EOF {
+				return err
+			}
+		}
+	}
+
+	err := p.shiftLines()
+	if err != nil && err != io.EOF {
+		return err
+	}
+
+	p.lineno++
+	if p.lines[0] == "" {
+		p.eof = true
+		return io.EOF
+	}
+	return nil
+}
+
+func (p *parser) shiftLines() (err error) {
+	for i := 0; i < len(p.lines)-1; i++ {
+		p.lines[i] = p.lines[i+1]
+	}
+	p.lines[len(p.lines)-1], err = p.r.ReadString('\n')
+	return
+}
+
+// Line returns a line from the parser without advancing it. A delta of 0
+// returns the current line, while higher deltas return read-ahead lines. It
+// returns an empty string if the delta is higher than the available lines,
+// either because of the buffer size or because the parser reached the end of
+// the input. Valid lines always contain at least a newline character.
+func (p *parser) Line(delta uint) string {
+	return p.lines[delta]
+}
+
+// Errorf generates an error and appends the current line information.
+func (p *parser) Errorf(delta int64, msg string, args ...interface{}) error {
+	return fmt.Errorf("gitdiff: line %d: %s", p.lineno+delta, fmt.Sprintf(msg, args...))
+}