Beispiel #1
0
// Removes the first parts of a string that match the pattern.
func removePattern(str string, pattern *regexp.Regexp) string {
	found := pattern.Find([]byte(str))

	if found != nil {
		return strings.Replace(str, string(found), "", 1)
	}

	return str
}
Beispiel #2
0
func (r *FetchResult) find(re *regexp.Regexp) (numbers []int) {
	numbers = make([]int, LOTTO_SIZE)
	lotto := re.Find(r.body)
	noExp := regexp.MustCompile(`\d+`)
	noBytes := noExp.FindAll(lotto, -1)
	for i := 0; i < len(noBytes); i++ {
		if no, err := strconv.Atoi(string(noBytes[i])); err == nil {
			numbers[i] = no
		}
	}
	return
}
Beispiel #3
0
// Search finds all lines matching regexp. context is the number of context
// lines to include before and after the match.
func (g *Grep) Search(r *regexp.Regexp, context int) []Match {
	matches := make([]Match, 0)

	for i, l := range g.lines {
		m := r.Find(l)
		if m == nil {
			continue
		}

		matches = append(matches, Match{
			LineNum:       i + 1,
			Match:         m,
			FullLine:      l,
			ContextBefore: g.context(i-context, i),
			ContextAfter:  g.context(i+1, i+1+context),
		})
	}

	return matches
}
Beispiel #4
0
func (s *SimpleScanner) Match(pattern string) ([]byte, Scanner) {
	var regc *regexp.Regexp
	var err error

	if pattern[0] != '^' {
		panic("match patterns must begin with `^`")
	}

	regc = s.patterns[pattern]
	if regc == nil {
		if regc, err = regexp.Compile(pattern); err == nil {
			s.patterns[pattern] = regc
		} else {
			panic(err.Error())
		}
	}
	if token := regc.Find(s.buf[s.cursor:]); token != nil {
		s.cursor += len(token)
		return token, s
	}
	return nil, s
}
Beispiel #5
0
func retrieve_urls(uri string, urlRegexp *regexp.Regexp) []string {
	urls := []string{}
	resp, err := http.Get(uri)
	if err != nil {
		return urls
	}
	defer resp.Body.Close()
	body, err := ioutil.ReadAll(resp.Body)
	lines := strings.Split(string(body), "\n")
	for _, v := range lines {
		matched, _ := regexp.MatchString("<a href=\"http", v)
		if matched {
			matchedurl := string(urlRegexp.Find([]byte(v)))
			matchedurl = fixUrl(matchedurl)
			visited[matchedurl]++
			if visited[matchedurl] > 1 {
				continue
			} else {
				urls = append(urls, matchedurl)
			}
		}
	}
	return urls
}
Beispiel #6
0
func (c *CssCompressor) performGeneralCleanup() {
	// This function does a lot, ok?
	var sb bytes.Buffer
	var previousIndex int
	var re *regexp.Regexp

	// Remove the spaces before the things that should not have spaces before them.
	// But, be careful not to turn "p :link {...}" into "p:link{...}"
	// Swap out any pseudo-class colons with the token, and then swap back.
	c.Css = RegexFindReplace(c.Css,
		"(^|\\})(([^\\{:])+:)+([^\\{]*\\{)",
		func(groups []string) string {
			s := groups[0]
			s = strings.Replace(s, ":", "___YUICSSMIN_PSEUDOCLASSCOLON___", -1)
			s = strings.Replace(s, "\\\\", "\\\\\\\\", -1)
			s = strings.Replace(s, "\\$", "\\\\\\$", -1)
			return s
		})

	// Remove spaces before the things that should not have spaces before them.
	re, _ = regexp.Compile("\\s+([!{};:>+\\(\\)\\],])")
	c.Css = re.ReplaceAll(c.Css, []byte("$1"))
	// Restore spaces for !important
	c.Css = bytes.Replace(c.Css, []byte("!important"), []byte(" !important"), -1)
	// bring back the colon
	c.Css = bytes.Replace(c.Css, []byte("___YUICSSMIN_PSEUDOCLASSCOLON___"), []byte(":"), -1)

	// retain space for special IE6 cases
	c.Css = RegexFindReplace(c.Css,
		"(?i):first\\-(line|letter)(\\{|,)",
		func(groups []string) string {
			return strings.ToLower(":first-"+groups[1]) + " " + groups[2]
		})

	// no space after the end of a preserved comment
	c.Css = bytes.Replace(c.Css, []byte("*/ "), []byte("*/"), -1)

	// If there are multiple @charset directives, push them to the top of the file.
	c.Css = RegexFindReplace(c.Css,
		"(?i)^(.*)(@charset)( \"[^\"]*\";)",
		func(groups []string) string {
			return strings.ToLower(groups[2]) + groups[3] + groups[1]
		})

	// When all @charset are at the top, remove the second and after (as they are completely ignored).
	c.Css = RegexFindReplace(c.Css,
		"(?i)^((\\s*)(@charset)( [^;]+;\\s*))+",
		func(groups []string) string {
			return groups[2] + strings.ToLower(groups[3]) + groups[4]
		})

	// lowercase some popular @directives
	c.Css = RegexFindReplace(c.Css,
		"(?i)@(charset|font-face|import|(?:-(?:atsc|khtml|moz|ms|o|wap|webkit)-)?keyframe|media|page|namespace)",
		func(groups []string) string {
			return "@" + strings.ToLower(groups[1])
		})

	// lowercase some more common pseudo-elements
	c.Css = RegexFindReplace(c.Css,
		"(?i):(active|after|before|checked|disabled|empty|enabled|first-(?:child|of-type)|focus|hover|last-(?:child|of-type)|link|only-(?:child|of-type)|root|:selection|target|visited)",
		func(groups []string) string {
			return ":" + strings.ToLower(groups[1])
		})

	// lowercase some more common functions
	c.Css = RegexFindReplace(c.Css,
		"(?i):(lang|not|nth-child|nth-last-child|nth-last-of-type|nth-of-type|(?:-(?:moz|webkit)-)?any)\\(",
		func(groups []string) string {
			return ":" + strings.ToLower(groups[1]) + "("
		})

	// lower case some common function that can be values
	// NOTE: rgb() isn't useful as we replace with #hex later, as well as and() is already done for us right after this
	c.Css = RegexFindReplace(c.Css,
		"(?i)([:,\\( ]\\s*)(attr|color-stop|from|rgba|to|url|(?:-(?:atsc|khtml|moz|ms|o|wap|webkit)-)?(?:calc|max|min|(?:repeating-)?(?:linear|radial)-gradient)|-webkit-gradient)",
		func(groups []string) string {
			return groups[1] + strings.ToLower(groups[2])
		})

	// Put the space back in some cases, to support stuff like
	// @media screen and (-webkit-min-device-pixel-ratio:0){
	re, _ = regexp.Compile("(?i)\\band\\(")
	c.Css = re.ReplaceAll(c.Css, []byte("and ("))

	// Remove the spaces after the things that should not have spaces after them.
	re, _ = regexp.Compile("([!{}:;>+\\(\\[,])\\s+")
	c.Css = re.ReplaceAll(c.Css, []byte("$1"))

	// remove unnecessary semicolons
	re, _ = regexp.Compile(";+}")
	c.Css = re.ReplaceAll(c.Css, []byte("}"))

	// Replace 0(px,em,%) with 0.
	re, _ = regexp.Compile("(?i)(^|[^0-9])(?:0?\\.)?0(?:px|em|%|in|cm|mm|pc|pt|ex|deg|g?rad|m?s|k?hz)")
	c.Css = re.ReplaceAll(c.Css, []byte("${1}0"))

	// Replace 0 0 0 0; with 0.
	re, _ = regexp.Compile(":0 0 0 0(;|})")
	re2, _ := regexp.Compile(":0 0 0(;|})")
	re3, _ := regexp.Compile(":0 0(;|})")
	c.Css = re.ReplaceAll(c.Css, []byte(":0$1"))
	c.Css = re2.ReplaceAll(c.Css, []byte(":0$1"))
	c.Css = re3.ReplaceAll(c.Css, []byte(":0$1"))

	// Replace background-position:0; with background-position:0 0;
	// same for transform-origin
	c.Css = RegexFindReplace(c.Css,
		"(?i)(background-position|webkit-mask-position|transform-origin|webkit-transform-origin|moz-transform-origin|o-transform-origin|ms-transform-origin):0(;|})",
		func(groups []string) string {
			return strings.ToLower(groups[1]) + ":0 0" + groups[2]
		})

	// Replace 0.6 to .6, but only when preceded by : or a white-space
	re, _ = regexp.Compile("(:|\\s)0+\\.(\\d+)")
	c.Css = re.ReplaceAll(c.Css, []byte("$1.$2"))

	// Shorten colors from rgb(51,102,153) to #336699
	// This makes it more likely that it'll get further compressed in the next step.
	c.Css = RegexFindReplace(c.Css,
		"rgb\\s*\\(\\s*([0-9,\\s]+)\\s*\\)",
		func(groups []string) string {
			rgbcolors := strings.Split(groups[1], ",")
			var hexcolor bytes.Buffer
			hexcolor.WriteString("#")
			for _, colour := range rgbcolors {
				val, _ := strconv.Atoi(colour)
				if val < 16 {
					hexcolor.WriteString("0")
				}
				// If someone passes an RGB value that's too big to express in two characters, round down.
				// Probably should throw out a warning here, but generating valid CSS is a bigger concern.
				if val > 255 {
					val = 255
				}
				hexcolor.WriteString(fmt.Sprintf("%x", val))
			}
			return hexcolor.String()
		})

	// Shorten colors from #AABBCC to #ABC. Note that we want to make sure
	// the color is not preceded by either ", " or =. Indeed, the property
	//     filter: chroma(color="#FFFFFF");
	// would become
	//     filter: chroma(color="#FFF");
	// which makes the filter break in IE.
	// We also want to make sure we're only compressing #AABBCC patterns inside { }, not id selectors ( #FAABAC {} )
	// We also want to avoid compressing invalid values (e.g. #AABBCCD to #ABCD)
	sb.Reset()
	re, _ = regexp.Compile("(\\=\\s*?[\"']?)?" + "#([0-9a-fA-F])([0-9a-fA-F])([0-9a-fA-F])([0-9a-fA-F])([0-9a-fA-F])([0-9a-fA-F])" + "(:?\\}|[^0-9a-fA-F{][^{]*?\\})")
	previousIndex = 0

	for match := re.Find(c.Css[previousIndex:]); match != nil; match = re.Find(c.Css[previousIndex:]) {
		index := re.FindIndex(c.Css[previousIndex:])
		submatches := re.FindStringSubmatch(string(c.Css[previousIndex:]))
		submatchIndexes := re.FindSubmatchIndex(c.Css[previousIndex:])

		sb.WriteString(string(c.Css[previousIndex : index[0]+len(c.Css[:previousIndex])]))

		//boolean isFilter = (m.group(1) != null && !"".equals(m.group(1)));
		// I hope the below is the equivalent of the above :P
		isFilter := submatches[1] != "" && submatchIndexes[1] != -1

		if isFilter {
			// Restore, as is. Compression will break filters
			sb.WriteString(submatches[1] + "#" + submatches[2] + submatches[3] + submatches[4] + submatches[5] + submatches[6] + submatches[7])
		} else {
			if strings.ToLower(submatches[2]) == strings.ToLower(submatches[3]) &&
				strings.ToLower(submatches[4]) == strings.ToLower(submatches[5]) &&
				strings.ToLower(submatches[6]) == strings.ToLower(submatches[7]) {
				// #AABBCC pattern
				sb.WriteString("#" + strings.ToLower(submatches[3]+submatches[5]+submatches[7]))
			} else {
				// Non-compressible color, restore, but lower case.
				sb.WriteString("#" + strings.ToLower(submatches[2]+submatches[3]+submatches[4]+submatches[5]+submatches[6]+submatches[7]))
			}
		}

		// The "+ 4" below is a crazy hack which will come back to haunt me later.
		// For now, it makes everything work 100%.
		previousIndex = submatchIndexes[7] + len(c.Css[:previousIndex]) + 4

	}
	if previousIndex > 0 {
		sb.WriteString(string(c.Css[previousIndex:]))
	}
	if sb.Len() > 0 {
		c.Css = sb.Bytes()
	}

	// Save a few chars by utilizing short colour keywords.
	// https://github.com/yui/yuicompressor/commit/fe8cf35d3693910103d65bf465d33b0d602dcfea
	colours := map[string]string{
		"#f00":    "red",
		"#000080": "navy",
		"#808080": "gray",
		"#808000": "olive",
		"#800080": "purple",
		"#c0c0c0": "silver",
		"#008080": "teal",
		"#ffa500": "orange",
		"#800000": "maroon",
	}
	for k, v := range colours {
		re, _ = regexp.Compile("(:|\\s)" + k + "(;|})")
		c.Css = re.ReplaceAll(c.Css, []byte("${1}"+v+"${2}"))
	}

	// border: none -> border:0
	c.Css = RegexFindReplace(c.Css,
		"(?i)(border|border-top|border-right|border-bottom|border-left|outline|background):none(;|})",
		func(groups []string) string {
			return strings.ToLower(groups[1]) + ":0" + groups[2]
		})

	// shorter opacity IE filter
	re, _ = regexp.Compile("(?i)progid:DXImageTransform.Microsoft.Alpha\\(Opacity=")
	c.Css = re.ReplaceAll(c.Css, []byte("alpha(opacity="))

	// Find a fraction that is used for Opera's -o-device-pixel-ratio query
	// Add token to add the "\" back in later
	re, _ = regexp.Compile("\\(([\\-A-Za-z]+):([0-9]+)\\/([0-9]+)\\)")
	c.Css = re.ReplaceAll(c.Css, []byte("(${1}:${2}___YUI_QUERY_FRACTION___${3})"))

	// Remove empty rules.
	re, _ = regexp.Compile("[^\\}\\{/;]+\\{\\}")
	c.Css = re.ReplaceAll(c.Css, []byte(""))

	// Add "\" back to fix Opera -o-device-pixel-ratio query
	c.Css = bytes.Replace(c.Css, []byte("___YUI_QUERY_FRACTION___"), []byte("/"), -1)
}