Add support for SubRip (#232)

* Add support for SubRip file format * Handle Windows newlines In Windows, a newline is represented as `\r\n`. * Add benchmark for regex vs time.Parse .srt detection * Make Srt timestamp detection more restrictive * Remove unused Srt functions * Use a constant for Srt timestamp layout * Remove redundant comments in Srt * make `Srt` work with Go <1.17 * Ensure correct Srt timestamp length before processing Co-authored-by: Gabriel Vasile <gabriel.vasile@email.com> * Remove `application/srt` alias from Srt Co-authored-by: Gabriel Vasile <gabriel.vasile@email.com>
gabriel-vasile · Jan 17, 2022 · e7d7726 · e7d7726
1 parent db2464c
commit e7d7726
Show file tree

Hide file tree

Showing 8 changed files with 177 additions and 45 deletions.
diff --git a/internal/magic/magic_test.go b/internal/magic/magic_test.go
@@ -112,3 +112,15 @@ func TestDropLastLine(t *testing.T) {
  }
  }
 }
+
+func BenchmarkSrt(b *testing.B) {
+ const subtitle = `1
+00:02:16,612 --> 00:02:19,376
+Senator, we're making
+our final approach into Coruscant.
+
+`
+ for i := 0; i < b.N; i++ {
+ Srt([]byte(subtitle), 0)
+ }
+}
diff --git a/internal/magic/text.go b/internal/magic/text.go
@@ -3,6 +3,8 @@ package magic
 import (
  "bufio"
  "bytes"
+ "strings"
+ "time"
 
  "github.com/gabriel-vasile/mimetype/internal/charset"
  "github.com/gabriel-vasile/mimetype/internal/json"
@@ -297,3 +299,52 @@ func HAR(raw []byte, limit uint32) bool {
 func Svg(raw []byte, limit uint32) bool {
  return bytes.Contains(raw, []byte("<svg"))
 }
+
+// Srt matches a SubRip file.
+func Srt(in []byte, _ uint32) bool {
+ s := bufio.NewScanner(bytes.NewReader(in))
+ if !s.Scan() {
+ return false
+ }
+ // First line must be 1.
+ if s.Text() != "1" {
+ return false
+ }
+
+ if !s.Scan() {
+ return false
+ }
+ secondLine := s.Text()
+ // Timestamp format (e.g: 00:02:16,612 --> 00:02:19,376) limits secondLine
+ // length to exactly 29 characters.
+ if len(secondLine) != 29 {
+ return false
+ }
+ // Decimal separator of fractional seconds in the timestamps must be a
+ // comma, not a period.
+ if strings.Contains(secondLine, ".") {
+ return false
+ }
+ // For Go <1.17, comma is not recognised as a decimal separator by `time.Parse`.
+ secondLine = strings.ReplaceAll(secondLine, ",", ".")
+ // Second line must be a time range.
+ ts := strings.Split(secondLine, " --> ")
+ if len(ts) != 2 {
+ return false
+ }
+ const layout = "15:04:05.000"
+ t0, err := time.Parse(layout, ts[0])
+ if err != nil {
+ return false
+ }
+ t1, err := time.Parse(layout, ts[1])
+ if err != nil {
+ return false
+ }
+ if t0.After(t1) {
+ return false
+ }
+
+ // A third line must exist and not be empty. This is the actual subtitle text.
+ return s.Scan() && len(s.Bytes()) != 0
+}
diff --git a/mimetype_test.go b/mimetype_test.go
@@ -171,47 +171,53 @@ var files = map[string]string{
  "shx.shx": "application/octet-stream",
  "so.so": "application/x-sharedlib",
  "sqlite.sqlite": "application/vnd.sqlite3",
- "svg.1.svg": "image/svg+xml",
- "svg.svg": "image/svg+xml",
- "swf.swf": "application/x-shockwave-flash",
- "tar.tar": "application/x-tar",
- "tcl.tcl": "text/x-tcl",
- "tcx.tcx": "application/vnd.garmin.tcx+xml",
- "tiff.tiff": "image/tiff",
- "torrent.torrent": "application/x-bittorrent",
- "tsv.tsv": "text/tab-separated-values",
- "ttf.ttf": "font/ttf",
- "tzfile": "application/tzif",
- "utf16bebom.txt": "text/plain; charset=utf-16be",
- "utf16lebom.txt": "text/plain; charset=utf-16le",
- "utf32bebom.txt": "text/plain; charset=utf-32be",
- "utf32lebom.txt": "text/plain; charset=utf-32le",
- "utf8.txt": "text/plain; charset=utf-8",
- "utf8ctrlchars": "application/octet-stream",
- "vcf.dos.vcf": "text/vcard",
- "vcf.vcf": "text/vcard",
- "voc.voc": "audio/x-unknown",
- "warc.warc": "application/warc",
- "wasm.wasm": "application/wasm",
- "wav.wav": "audio/wav",
- "webm.webm": "video/webm",
- "webp.webp": "image/webp",
- "woff.woff": "font/woff",
- "woff2.woff2": "font/woff2",
- "x3d.x3d": "model/x3d+xml",
- "xar.xar": "application/x-xar",
- "xcf.xcf": "image/x-xcf",
- "xfdf.xfdf": "application/vnd.adobe.xfdf",
- "xlf.xlf": "application/x-xliff+xml",
- "xls.xls": "application/vnd.ms-excel",
- "xlsx.1.xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
- "xlsx.2.xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
- "xlsx.xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
- "xml.xml": "text/xml; charset=utf-8",
- "xml.withbr.xml": "text/xml; charset=utf-8",
- "xz.xz": "application/x-xz",
- "zip.zip": "application/zip",
- "zst.zst": "application/zstd",
+ "srt.srt": "text/x-subrip",
+ // not.srt.txt uses periods instead of commas for the decimal separators of
+ // the timestamps.
+ "not.srt.txt": "text/plain; charset=utf-8",
+ // not.srt.2.txt does not specify milliseconds.
+ "not.srt.2.txt": "text/plain; charset=utf-8",
+ "svg.1.svg": "image/svg+xml",
+ "svg.svg": "image/svg+xml",
+ "swf.swf": "application/x-shockwave-flash",
+ "tar.tar": "application/x-tar",
+ "tcl.tcl": "text/x-tcl",
+ "tcx.tcx": "application/vnd.garmin.tcx+xml",
+ "tiff.tiff": "image/tiff",
+ "torrent.torrent": "application/x-bittorrent",
+ "tsv.tsv": "text/tab-separated-values",
+ "ttf.ttf": "font/ttf",
+ "tzfile": "application/tzif",
+ "utf16bebom.txt": "text/plain; charset=utf-16be",
+ "utf16lebom.txt": "text/plain; charset=utf-16le",
+ "utf32bebom.txt": "text/plain; charset=utf-32be",
+ "utf32lebom.txt": "text/plain; charset=utf-32le",
+ "utf8.txt": "text/plain; charset=utf-8",
+ "utf8ctrlchars": "application/octet-stream",
+ "vcf.dos.vcf": "text/vcard",
+ "vcf.vcf": "text/vcard",
+ "voc.voc": "audio/x-unknown",
+ "warc.warc": "application/warc",
+ "wasm.wasm": "application/wasm",
+ "wav.wav": "audio/wav",
+ "webm.webm": "video/webm",
+ "webp.webp": "image/webp",
+ "woff.woff": "font/woff",
+ "woff2.woff2": "font/woff2",
+ "x3d.x3d": "model/x3d+xml",
+ "xar.xar": "application/x-xar",
+ "xcf.xcf": "image/x-xcf",
+ "xfdf.xfdf": "application/vnd.adobe.xfdf",
+ "xlf.xlf": "application/x-xliff+xml",
+ "xls.xls": "application/vnd.ms-excel",
+ "xlsx.1.xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
+ "xlsx.2.xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
+ "xlsx.xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
+ "xml.xml": "text/xml; charset=utf-8",
+ "xml.withbr.xml": "text/xml; charset=utf-8",
+ "xz.xz": "application/x-xz",
+ "zip.zip": "application/zip",
+ "zst.zst": "application/zstd",
 }
 
 func TestDetect(t *testing.T) {

diff --git a/supported_mimes.md b/supported_mimes.md
@@ -1,4 +1,4 @@
-## 166 Supported MIME types
+## 167 Supported MIME types
 This file is automatically generated when running tests. Do not edit manually.
 
 Extension | MIME type | Aliases
@@ -163,6 +163,7 @@ Extension | MIME type | Aliases
 **.har** | application/json | -
 **.ndjson** | application/x-ndjson | -
 **.rtf** | text/rtf | -
+**.srt** | text/x-subrip | text/x-srt
 **.tcl** | text/x-tcl | application/x-tcl
 **.csv** | text/csv | -
 **.tsv** | text/tab-separated-values | -

diff --git a/testdata/not.srt.2.txt b/testdata/not.srt.2.txt
@@ -0,0 +1,20 @@
+1
+00:02:16 --> 00:02:19
+Senator, we're making
+our final approach into Coruscant.
+
+2
+00:02:19 --> 00:02:21
+Very good, Lieutenant.
+
+3
+00:03:13 --> 00:03:15
+We made it.
+
+4
+00:03:18 --> 00:03:20
+I guess I was wrong.
+
+5
+00:03:20 --> 00:03:22
+There was no danger at all.
diff --git a/testdata/not.srt.txt b/testdata/not.srt.txt
@@ -0,0 +1,20 @@
+1
+00:02:16.612 --> 00:02:19.376
+Senator, we're making
+our final approach into Coruscant.
+
+2
+00:02:19.482 --> 00:02:21.609
+Very good, Lieutenant.
+
+3
+00:03:13.336 --> 00:03:15.167
+We made it.
+
+4
+00:03:18.608 --> 00:03:20.371
+I guess I was wrong.
+
+5
+00:03:20.476 --> 00:03:22.671
+There was no danger at all.
diff --git a/testdata/srt.srt b/testdata/srt.srt
@@ -0,0 +1,20 @@
+1
+00:02:16,612 --> 00:02:19,376
+Senator, we're making
+our final approach into Coruscant.
+
+2
+00:02:19,482 --> 00:02:21,609
+Very good, Lieutenant.
+
+3
+00:03:13,336 --> 00:03:15,167
+We made it.
+
+4
+00:03:18,608 --> 00:03:20,371
+I guess I was wrong.
+
+5
+00:03:20,476 --> 00:03:22,671
+There was no danger at all.
diff --git a/tree.go b/tree.go
@@ -76,7 +76,7 @@ var (
  alias("application/x-ogg")
  oggAudio = newMIME("audio/ogg", ".oga", magic.OggAudio)
  oggVideo = newMIME("video/ogg", ".ogv", magic.OggVideo)
- text = newMIME("text/plain", ".txt", magic.Text, html, svg, xml, php, js, lua, perl, python, json, ndJSON, rtf, tcl, csv, tsv, vCard, iCalendar, warc)
+ text = newMIME("text/plain", ".txt", magic.Text, html, svg, xml, php, js, lua, perl, python, json, ndJSON, rtf, srt, tcl, csv, tsv, vCard, iCalendar, warc)
  xml = newMIME("text/xml", ".xml", magic.XML, rss, atom, x3d, kml, xliff, collada, gml, gpx, tcx, amf, threemf, xfdf, owl2)
  json = newMIME("application/json", ".json", magic.JSON, geoJSON, har)
  har = newMIME("application/json", ".har", magic.HAR)
@@ -87,8 +87,10 @@ var (
  html = newMIME("text/html", ".html", magic.HTML)
  php = newMIME("text/x-php", ".php", magic.Php)
  rtf = newMIME("text/rtf", ".rtf", magic.Rtf)
- js = newMIME("application/javascript", ".js", magic.Js).
- alias("application/x-javascript", "text/javascript")
+ srt = newMIME("text/x-subrip", ".srt", magic.Srt).
+ alias("text/x-srt")
+ js = newMIME("application/javascript", ".js", magic.Js).
+ alias("application/x-javascript", "text/javascript")
  lua = newMIME("text/x-lua", ".lua", magic.Lua)
  perl = newMIME("text/x-perl", ".pl", magic.Perl)
  python = newMIME("application/x-python", ".py", magic.Python)