Add more syntax files and include syntax highlighter in the repo

cmd/micro/buffer.go

@@ -16,7 +16,7 @@ import (
 	"unicode/utf8"
 
 	"github.com/mitchellh/go-homedir"
-	"github.com/zyedidia/highlight"
+	"github.com/zyedidia/micro/cmd/micro/highlight"
 )
 
 // Buffer stores the text for files that are loaded into the text editor

cmd/micro/highlight/ftdetect.go

@@ -0,0 +1,19 @@
+package highlight
+
+func DetectFiletype(defs []*Def, filename string, firstLine []byte) *Def {
+	for _, d := range defs {
+		if d.ftdetect[0].Match([]byte(filename)) {
+			return d
+		}
+		if len(d.ftdetect) > 1 {
+			if d.ftdetect[1].Match(firstLine) {
+				return d
+			}
+		}
+	}
+
+	emptyDef := new(Def)
+	emptyDef.FileType = "Unknown"
+	emptyDef.rules = new(Rules)
+	return emptyDef
+}

cmd/micro/highlight/highlighter.go

@@ -0,0 +1,262 @@
+package highlight
+
+import (
+	"regexp"
+	"strings"
+)
+
+func combineLineMatch(src, dst LineMatch) LineMatch {
+	for k, v := range src {
+		if g, ok := dst[k]; ok {
+			if g == "" {
+				dst[k] = v
+			}
+		} else {
+			dst[k] = v
+		}
+	}
+	return dst
+}
+
+type State *Region
+
+type LineStates interface {
+	LineData() [][]byte
+	State(lineN int) State
+	SetState(lineN int, s State)
+	SetMatch(lineN int, m LineMatch)
+}
+
+type Highlighter struct {
+	lastRegion *Region
+	def        *Def
+}
+
+func NewHighlighter(def *Def) *Highlighter {
+	h := new(Highlighter)
+	h.def = def
+	return h
+}
+
+type LineMatch map[int]string
+
+func FindIndex(regex *regexp.Regexp, str []byte, canMatchStart, canMatchEnd bool) []int {
+	regexStr := regex.String()
+	if strings.Contains(regexStr, "^") {
+		if !canMatchStart {
+			return nil
+		}
+	}
+	if strings.Contains(regexStr, "$") {
+		if !canMatchEnd {
+			return nil
+		}
+	}
+	return regex.FindIndex(str)
+}
+
+func FindAllIndex(regex *regexp.Regexp, str []byte, canMatchStart, canMatchEnd bool) [][]int {
+	regexStr := regex.String()
+	if strings.Contains(regexStr, "^") {
+		if !canMatchStart {
+			return nil
+		}
+	}
+	if strings.Contains(regexStr, "$") {
+		if !canMatchEnd {
+			return nil
+		}
+	}
+	return regex.FindAllIndex(str, -1)
+}
+
+func (h *Highlighter) highlightRegion(start int, canMatchEnd bool, lineNum int, line []byte, region *Region) LineMatch {
+	highlights := make(LineMatch)
+
+	if len(line) == 0 {
+		if canMatchEnd {
+			h.lastRegion = region
+		}
+
+		return highlights
+	}
+
+	loc := FindIndex(region.end, line, start == 0, canMatchEnd)
+	if loc != nil {
+		if region.parent == nil {
+			highlights[start+loc[1]] = ""
+			return combineLineMatch(highlights,
+				combineLineMatch(h.highlightRegion(start, false, lineNum, line[:loc[0]], region),
+					h.highlightEmptyRegion(start+loc[1], canMatchEnd, lineNum, line[loc[1]:])))
+		}
+		highlights[start+loc[1]] = region.parent.group
+		return combineLineMatch(highlights,
+			combineLineMatch(h.highlightRegion(start, false, lineNum, line[:loc[0]], region),
+				h.highlightRegion(start+loc[1], canMatchEnd, lineNum, line[loc[1]:], region.parent)))
+	}
+
+	firstLoc := []int{len(line), 0}
+	var firstRegion *Region
+	for _, r := range region.rules.regions {
+		loc := FindIndex(r.start, line, start == 0, canMatchEnd)
+		if loc != nil {
+			if loc[0] < firstLoc[0] {
+				firstLoc = loc
+				firstRegion = r
+			}
+		}
+	}
+	if firstLoc[0] != len(line) {
+		highlights[start+firstLoc[0]] = firstRegion.group
+		return combineLineMatch(highlights,
+			combineLineMatch(h.highlightRegion(start, false, lineNum, line[:firstLoc[0]], region),
+				h.highlightRegion(start+firstLoc[1], canMatchEnd, lineNum, line[firstLoc[1]:], firstRegion)))
+	}
+
+	for _, p := range region.rules.patterns {
+		matches := FindAllIndex(p.regex, line, start == 0, canMatchEnd)
+		for _, m := range matches {
+			highlights[start+m[0]] = p.group
+			if _, ok := highlights[start+m[1]]; !ok {
+				highlights[start+m[1]] = region.group
+			}
+		}
+	}
+
+	if canMatchEnd {
+		h.lastRegion = region
+	}
+
+	return highlights
+}
+
+func (h *Highlighter) highlightEmptyRegion(start int, canMatchEnd bool, lineNum int, line []byte) LineMatch {
+	highlights := make(LineMatch)
+	if len(line) == 0 {
+		if canMatchEnd {
+			h.lastRegion = nil
+		}
+		return highlights
+	}
+
+	firstLoc := []int{len(line), 0}
+	var firstRegion *Region
+	for _, r := range h.def.rules.regions {
+		loc := FindIndex(r.start, line, start == 0, canMatchEnd)
+		if loc != nil {
+			if loc[0] < firstLoc[0] {
+				firstLoc = loc
+				firstRegion = r
+			}
+		}
+	}
+	if firstLoc[0] != len(line) {
+		highlights[start+firstLoc[0]] = firstRegion.group
+		return combineLineMatch(highlights,
+			combineLineMatch(h.highlightEmptyRegion(start, false, lineNum, line[:firstLoc[0]]),
+				h.highlightRegion(start+firstLoc[1], canMatchEnd, lineNum, line[firstLoc[1]:], firstRegion)))
+	}
+
+	for _, p := range h.def.rules.patterns {
+		matches := FindAllIndex(p.regex, line, start == 0, canMatchEnd)
+		for _, m := range matches {
+			highlights[start+m[0]] = p.group
+			if _, ok := highlights[start+m[1]]; !ok {
+				highlights[start+m[1]] = ""
+			}
+		}
+	}
+
+	if canMatchEnd {
+		h.lastRegion = nil
+	}
+
+	return highlights
+}
+
+func (h *Highlighter) HighlightString(input string) []LineMatch {
+	lines := strings.Split(input, "\n")
+	var lineMatches []LineMatch
+
+	for i := 0; i < len(lines); i++ {
+		line := []byte(lines[i])
+
+		if i == 0 || h.lastRegion == nil {
+			lineMatches = append(lineMatches, h.highlightEmptyRegion(0, true, i, line))
+		} else {
+			lineMatches = append(lineMatches, h.highlightRegion(0, true, i, line, h.lastRegion))
+		}
+	}
+
+	return lineMatches
+}
+
+func (h *Highlighter) Highlight(input LineStates, startline int) {
+	lines := input.LineData()
+
+	for i := startline; i < len(lines); i++ {
+		line := []byte(lines[i])
+
+		var match LineMatch
+		if i == 0 || h.lastRegion == nil {
+			match = h.highlightEmptyRegion(0, true, i, line)
+		} else {
+			match = h.highlightRegion(0, true, i, line, h.lastRegion)
+		}
+
+		curState := h.lastRegion
+
+		input.SetMatch(i, match)
+		input.SetState(i, curState)
+	}
+}
+
+func (h *Highlighter) ReHighlightLine(input LineStates, lineN int) {
+	lines := input.LineData()
+
+	line := []byte(lines[lineN])
+
+	h.lastRegion = nil
+	if lineN > 0 {
+		h.lastRegion = input.State(lineN - 1)
+	}
+
+	var match LineMatch
+	if lineN == 0 || h.lastRegion == nil {
+		match = h.highlightEmptyRegion(0, true, lineN, line)
+	} else {
+		match = h.highlightRegion(0, true, lineN, line, h.lastRegion)
+	}
+	curState := h.lastRegion
+
+	input.SetMatch(lineN, match)
+	input.SetState(lineN, curState)
+}
+
+func (h *Highlighter) ReHighlight(input LineStates, startline int) {
+	lines := input.LineData()
+
+	h.lastRegion = nil
+	if startline > 0 {
+		h.lastRegion = input.State(startline - 1)
+	}
+	for i := startline; i < len(lines); i++ {
+		line := []byte(lines[i])
+
+		var match LineMatch
+		if i == 0 || h.lastRegion == nil {
+			match = h.highlightEmptyRegion(0, true, i, line)
+		} else {
+			match = h.highlightRegion(0, true, i, line, h.lastRegion)
+		}
+		curState := h.lastRegion
+		lastState := input.State(i)
+
+		input.SetMatch(i, match)
+		input.SetState(i, curState)
+
+		if curState == lastState {
+			break
+		}
+	}
+}

cmd/micro/highlight/parser.go

@@ -0,0 +1,202 @@
+package highlight
+
+import (
+	"fmt"
+	"regexp"
+
+	"gopkg.in/yaml.v2"
+)
+
+// A Def is a full syntax definition for a language
+// It has a filetype, information about how to detect the filetype based
+// on filename or header (the first line of the file)
+// Then it has the rules which define how to highlight the file
+type Def struct {
+	FileType string
+	ftdetect []*regexp.Regexp
+	rules    *Rules
+}
+
+// A Pattern is one simple syntax rule
+// It has a group that the rule belongs to, as well as
+// the regular expression to match the pattern
+type Pattern struct {
+	group string
+	regex *regexp.Regexp
+}
+
+// Rules defines which patterns and regions can be used to highlight
+// a filetype
+type Rules struct {
+	regions  []*Region
+	patterns []*Pattern
+	includes []string
+}
+
+// A Region is a highlighted region (such as a multiline comment, or a string)
+// It belongs to a group, and has start and end regular expressions
+// A Region also has rules of its own that only apply when matching inside the
+// region and also rules from the above region do not match inside this region
+// Note that a region may contain more regions
+type Region struct {
+	group  string
+	parent *Region
+	start  *regexp.Regexp
+	end    *regexp.Regexp
+	rules  *Rules
+}
+
+// ParseDef parses an input syntax file into a highlight Def
+func ParseDef(input []byte) (s *Def, err error) {
+	// This is just so if we have an error, we can exit cleanly and return the parse error to the user
+	defer func() {
+		if e := recover(); e != nil {
+			err = e.(error)
+		}
+	}()
+
+	var rules map[interface{}]interface{}
+	if err = yaml.Unmarshal(input, &rules); err != nil {
+		return nil, err
+	}
+
+	s = new(Def)
+
+	for k, v := range rules {
+		if k == "filetype" {
+			filetype := v.(string)
+
+			s.FileType = filetype
+		} else if k == "detect" {
+			ftdetect := v.(map[interface{}]interface{})
+			if len(ftdetect) >= 1 {
+				syntax, err := regexp.Compile(ftdetect["filename"].(string))
+				if err != nil {
+					return nil, err
+				}
+
+				s.ftdetect = append(s.ftdetect, syntax)
+			}
+			if len(ftdetect) >= 2 {
+				header, err := regexp.Compile(ftdetect["header"].(string))
+				if err != nil {
+					return nil, err
+				}
+
+				s.ftdetect = append(s.ftdetect, header)
+			}
+		} else if k == "rules" {
+			inputRules := v.([]interface{})
+
+			rules, err := parseRules(inputRules, nil)
+			if err != nil {
+				return nil, err
+			}
+
+			s.rules = rules
+		}
+	}
+
+	return s, err
+}
+
+func ResolveIncludes(defs []*Def) {
+	for _, d := range defs {
+		resolveIncludesInDef(defs, d)
+	}
+}
+
+func resolveIncludesInDef(defs []*Def, d *Def) {
+	for _, lang := range d.rules.includes {
+		for _, searchDef := range defs {
+			if lang == searchDef.FileType {
+				d.rules.patterns = append(d.rules.patterns, searchDef.rules.patterns...)
+				d.rules.regions = append(d.rules.regions, searchDef.rules.regions...)
+			}
+		}
+	}
+	for _, r := range d.rules.regions {
+		resolveIncludesInRegion(defs, r)
+		r.parent = nil
+	}
+}
+
+func resolveIncludesInRegion(defs []*Def, region *Region) {
+	for _, lang := range region.rules.includes {
+		for _, searchDef := range defs {
+			if lang == searchDef.FileType {
+				region.rules.patterns = append(region.rules.patterns, searchDef.rules.patterns...)
+				region.rules.regions = append(region.rules.regions, searchDef.rules.regions...)
+			}
+		}
+	}
+	for _, r := range region.rules.regions {
+		resolveIncludesInRegion(defs, r)
+		r.parent = region
+	}
+}
+
+func parseRules(input []interface{}, curRegion *Region) (*Rules, error) {
+	rules := new(Rules)
+
+	for _, v := range input {
+		rule := v.(map[interface{}]interface{})
+		for k, val := range rule {
+			group := k
+
+			switch object := val.(type) {
+			case string:
+				if k == "include" {
+					rules.includes = append(rules.includes, object)
+				} else {
+					// Pattern
+					r, err := regexp.Compile(object)
+					if err != nil {
+						return nil, err
+					}
+
+					rules.patterns = append(rules.patterns, &Pattern{group.(string), r})
+				}
+			case map[interface{}]interface{}:
+				// Region
+				region, err := parseRegion(group.(string), object, curRegion)
+				if err != nil {
+					return nil, err
+				}
+				rules.regions = append(rules.regions, region)
+			default:
+				return nil, fmt.Errorf("Bad type %T", object)
+			}
+		}
+	}
+
+	return rules, nil
+}
+
+func parseRegion(group string, regionInfo map[interface{}]interface{}, prevRegion *Region) (*Region, error) {
+	var err error
+
+	region := new(Region)
+	region.group = group
+	region.parent = prevRegion
+
+	region.start, err = regexp.Compile(regionInfo["start"].(string))
+
+	if err != nil {
+		return nil, err
+	}
+
+	region.end, err = regexp.Compile(regionInfo["end"].(string))
+
+	if err != nil {
+		return nil, err
+	}
+
+	region.rules, err = parseRules(regionInfo["rules"].([]interface{}), region)
+
+	if err != nil {
+		return nil, err
+	}
+
+	return region, nil
+}

cmd/micro/highlighter.go

@@ -1,6 +1,6 @@
 package main
 
-import "github.com/zyedidia/highlight"
+import "github.com/zyedidia/micro/cmd/micro/highlight"
 
 var syntaxDefs []*highlight.Def
 

cmd/micro/lineArray.go

@@ -5,7 +5,7 @@ import (
 	"io"
 	"unicode/utf8"
 
-	"github.com/zyedidia/highlight"
+	"github.com/zyedidia/micro/cmd/micro/highlight"
 )
 
 func runeToByteIndex(n int, txt []byte) int {

runtime/syntax/arduino.yaml

@@ -0,0 +1,98 @@
+filetype: ino
+
+detect:
+    filename: "\\.?ino$"
+
+rules:
+    - identifier: "\\b[A-Z_][0-9A-Z_]+\\b" 
+
+      ## 
+    - type: "\\b((s?size)|((u_?)?int(8|16|32|64|ptr)))_t\\b"
+
+      ## Constants
+    - constant: "(?i)\\b(HIGH|LOW|INPUT|OUTPUT)\\b"
+
+      ## Serial Print
+    - constant: "(?i)\\b(DEC|BIN|HEX|OCT|BYTE)\\b"
+
+      ## PI Constants
+    - constant: "(?i)\\b(PI|HALF_PI|TWO_PI)\\b"
+
+      ## ShiftOut
+    - constant: "(?i)\\b(LSBFIRST|MSBFIRST)\\b"
+
+      ## Attach Interrupt
+    - constant: "(?i)\\b(CHANGE|FALLING|RISING)\\b"
+
+      ## Analog Reference
+    - constant: "(?i)\\b(DEFAULT|EXTERNAL|INTERNAL|INTERNAL1V1|INTERNAL2V56)\\b"
+
+      ## === FUNCTIONS === ##
+
+      ## Data Types
+    - type: "\\b(boolean|byte|char|float|int|long|word)\\b"
+
+      ## Control Structions
+    - statement: "\\b(case|class|default|do|double|else|false|for|if|new|null|private|protected|public|short|signed|static|String|switch|this|throw|try|true|unsigned|void|while)\\b" 
+    - statement: "\\b(goto|continue|break|return)\\b"
+
+      ## Math
+    - identifier: "\\b(abs|acos|asin|atan|atan2|ceil|constrain|cos|degrees|exp|floor|log|map|max|min|radians|random|randomSeed|round|sin|sq|sqrt|tan)\\b"
+
+      ## Bits & Bytes
+    - identifier: "\\b(bitRead|bitWrite|bitSet|bitClear|bit|highByte|lowByte)\\b"
+
+      ## Analog I/O
+    - identifier: "\\b(analogReference|analogRead|analogWrite)\\b"
+
+      ## External Interrupts
+    - identifier: "\\b(attachInterrupt|detachInterrupt)\\b"
+
+      ## Time
+    - identifier: "\\b(delay|delayMicroseconds|millis|micros)\\b"
+
+      ## Digital I/O
+    - identifier: "\\b(pinMode|digitalWrite|digitalRead)\\b"
+
+      ## Interrupts
+    - identifier: "\\b(interrupts|noInterrupts)\\b"
+
+      ## Advanced I/O
+    - identifier: "\\b(noTone|pulseIn|shiftIn|shiftOut|tone)\\b"
+
+      ## Serial
+    - identifier: "\\b(Serial|Serial1|Serial2|Serial3|begin|end|peek|read|print|println|available|flush)\\b"
+
+      ## Structure
+    - identifier: "\\b(setup|loop)\\b"
+
+      ## 
+    - statement: "^[[:space:]]*#[[:space:]]*(define|include(_next)?|(un|ifn?)def|endif|el(if|se)|if|warning|error|pragma)"
+
+      ## GCC builtins
+    - constant: "(__attribute__[[:space:]]*\\(\\([^)]*\\)\\)|__(aligned|asm|builtin|hidden|inline|packed|restrict|section|typeof|weak)__)"
+
+    - constant.string:
+        start: "\""
+        end: "\""
+        rules:
+            - constant.specialChar: "\\\\."
+
+    - constant.string:
+        start: "'"
+        end: "'"
+        rules:
+            - preproc: "..+"
+            - constant.specialChar: "\\\\."
+
+    - comment:
+        start: "//"
+        end: "$"
+        rules:
+            - todo: "(TODO|XXX|FIXME):?"
+
+    - comment:
+        start: "/\\*"
+        end: "\\*/"
+        rules:
+            - todo: "(TODO|XXX|FIXME):?"

runtime/syntax/asm.yaml

@@ -0,0 +1,107 @@
+filetype: asm
+
+detect:
+    filename: "\\.(S|s|asm)$"
+
+rules:
+    # This file is made for NASM assembly
+
+    ## Instructions
+    # x86
+    - statement: "\\b(?i)(mov|aaa|aad|aam|aas|adc|add|and|call|cbw|clc|cld|cli|cmc|cmp|cmpsb|cmpsw|cwd|daa|das|dec|div|esc|hlt|idiv|imul|in|inc|int|into|iret|ja|jae|jb|jbe|jc|je|jg|jge|jl|jle|jna|jnae|jnb|jnbe|jnc|jne|jng|jnge|jnl|jnle|jno|jnp|jns|jnz|jo|jp|jpe|jpo|js|jz|jcxz|jmp|lahf|lds|lea|les|lock|lodsb|lodsw|loop|loope|loopne|loopnz|loopz|movsb|movsw|mul|neg|nop|or|pop|popf|push|pushf|rcl|rcr|rep|repe|repne|repnz|repz|ret|retn|retf|rol|ror|sahf|sal|sar|sbb|scasb|scasw|shl|shr|stc|std|sti|stosb|stosw|sub|test|wait|xchg|xlat|xor)(?-i)\\b"
+    - statement: "\\b(?i)(bound|enter|ins|leave|outs|popa|pusha)(?-i)\\b"
+    - statement: "\\b(?i)(arpl|clts|lar|lgdt|lidt|lldt|lmsw|loadall|lsl|ltr|sgdt|sidt|sldt|smsw|str|verr|verw)(?-i)\\b"
+    - statement: "\\b(?i)(bsf|bsr|bt|btc|btr|bts|cdq|cmpsd|cwde|insd|iret|iretd|iretf|jecxz|lfs|lgs|lss|lodsd|loopw|loopew|loopnew|loopnzw|loopzw|loopd|looped|loopned|loopnzd|loopzd|cr|tr|dr|movsd|movsx|movzx|outsd|popad|popfd|pushad|pushfd|scasd|seta|setae|setb|setbe|setc|sete|setg|setge|setl|setle|setna|setnae|setnb|setnbe|setnc|setne|setng|setnge|setnl|setnle|setno|setnp|setns|setnz|seto|setp|setpe|setpo|sets|setz|shdl|shrd|stosd)(?-i)\\b"
+    - statement: "\\b(?i)(bswap|cmpxcgh|invd|invlpg|wbinvd|xadd)(?-i)\\b"
+    - statement: "\\b(?i)(cpuid|cmpxchg8b|rdmsr|rdtsc|wrmsr|rsm)(?-i)\\b"
+    - statement: "\\b(?i)(rdpmc)(?-i)\\b"
+    - statement: "\\b(?i)(syscall|sysret)(?-i)\\b"
+    - statement: "\\b(?i)(cmova|cmovae|cmovb|cmovbe|cmovc|cmove|cmovg|cmovge|cmovl|cmovle|cmovna|cmovnae|cmovnb|cmovnbe|cmovnc|cmovne|cmovng|cmovnge|cmovnle|cmovno|cmovpn|cmovns|cmovnz|cmovo|cmovp|cmovpe|cmovpo|cmovs|cmovz|sysenter|sysexit|ud2)(?-i)\\b"
+    - statement: "\\b(?i)(maskmovq|movntps|movntq|prefetch0|prefetch1|prefetch2|prefetchnta|sfence)(?-i)\\b"
+    - statement: "\\b(?i)(clflush|lfence|maskmovdqu|mfence|movntdq|movnti|movntpd|pause)(?-i)\\b"
+    - statement: "\\b(?i)(monitor|mwait)(?-i)\\b"
+    - statement: "\\b(?i)(cdqe|cqo|cmpsq|cmpxchg16b|iretq|jrcxz|lodsq|movsdx|popfq|pushfq|rdtscp|scasq|stosq|swapgs)(?-i)\\b"
+    - statement: "\\b(?i)(clgi|invlpga|skinit|stgi|vmload|vmmcall|vmrun|vmsave)(?-i)\\b"
+    - statement: "\\b(?i)(vmptrdl|vmptrst|vmclear|vmread|vmwrite|vmcall|vmlaunch|vmresume|vmxoff|vmxon)(?-i)\\b"
+    - statement: "\\b(?i)(lzcnt|popcnt)(?-i)\\b"
+    - statement: "\\b(?i)(bextr|blcfill|blci|blcic|blcmask|blcs|blsfill|blsic|t1mskc|tzmsk)(?-i)\\b"
+
+      # x87
+    - statement: "\\b(?i)(f2xm1|fabs|fadd|faddp|fbld|fbstp|fchs|fclex|fcom|fcomp|fcompp|fdecstp|fdisi|fdiv|fvidp|fdivr|fdivrp|feni|ffree|fiadd|ficom|ficomp|fidiv|fidivr|fild|fimul|fincstp|finit|fist|fistp|fisub|fisubr|fld|fld1|fldcw|fldenv|fldenvw|fldl2e|fldl2t|fldlg2|fldln2|fldpi|fldz|fmul|fmulp|fnclex|fndisi|fneni|fninit|fnop|fnsave|fnsavenew|fnstcw|fnstenv|fnstenvw|fnstsw|fpatan|fprem|fptan|frndint|frstor|frstorw|fsave|fsavew|fscale|fsqrt|fst|fstcw|fstenv|fstenvw|fstp|fstpsw|fsub|fsubp|fsubr|fsubrp|ftst|fwait|fxam|fxch|fxtract|fyl2x|fyl2xp1)(?-i)\\b"
+    - statement: "\\b(?i)(fsetpm)(?-i)\\b"
+    - statement: "\\b(?i)(fcos|fldenvd|fsaved|fstenvd|fprem1|frstord|fsin|fsincos|fstenvd|fucom|fucomp|fucompp)(?-i)\\b"
+    - statement: "\\b(?i)(fcmovb|fcmovbe|fcmove|fcmove|fcmovnb|fcmovnbe|fcmovne|fcmovnu|fcmovu)(?-i)\\b"
+    - statement: "\\b(?i)(fcomi|fcomip|fucomi|fucomip)(?-i)\\b"
+    - statement: "\\b(?i)(fxrstor|fxsave)(?-i)\\b"
+    - statement: "\\b(?i)(fisttp)(?-i)\\b"
+    - statement: "\\b(?i)(ffreep)(?-i)\\b"
+
+      # SIMD
+    - statement: "\\b(?i)(emms|movd|movq|packssdw|packsswb|packuswb|paddb|paddw|paddd|paddsb|paddsw|paddusb|paddusw|pand|pandn|por|pxor|pcmpeqb|pcmpeqw|pcmpeqd|pcmpgtb|pcmpgtw|pcmpgtd|pmaddwd|pmulhw|pmullw|psllw|pslld|psllq|psrad|psraw|psrlw|psrld|psrlq|psubb|psubw|psubd|psubsb|psubsw|psubusb|punpckhbw|punpckhwd|punpckhdq|punkcklbw|punpckldq|punpcklwd)(?-i)\\b"
+    - statement: "\\b(?i)(paveb|paddsiw|pmagw|pdistib|psubsiw|pmwzb|pmulhrw|pmvnzb|pmvlzb|pmvgezb|pmulhriw|pmachriw)(?-i)\\b"
+    - statement: "\\b(?i)(femms|pavgusb|pf2id|pfacc|pfadd|pfcmpeq|pfcmpge|pfcmpgt|pfmax|pfmin|pfmul|pfrcp|pfrcpit1|pfrcpit2|pfrsqit1|pfrsqrt|pfsub|pfsubr|pi2fd|pmulhrw|prefetch|prefetchw)(?-i)\\b"
+    - statement: "\\b(?i)(pf2iw|pfnacc|pfpnacc|pi2fw|pswapd)(?-i)\\b"
+    - statement: "\\b(?i)(pfrsqrtv|pfrcpv)(?-i)\\b"
+    - statement: "\\b(?i)(addps|addss|cmpps|cmpss|comiss|cvtpi2ps|cvtps2pi|cvtsi2ss|cvtss2si|cvttps2pi|cvttss2si|divps|divss|ldmxcsr|maxps|maxss|minps|minss|movaps|movhlps|movhps|movlhps|movlps|movmskps|movntps|movss|movups|mulps|mulss|rcpps|rcpss|rsqrtps|rsqrtss|shufps|sqrtps|sqrtss|stmxcsr|subps|subss|ucomiss|unpckhps|unpcklps)(?-i)\\b"
+    - statement: "\\b(?i)(andnps|andps|orps|pavgb|pavgw|pextrw|pinsrw|pmaxsw|pmaxub|pminsw|pminub|pmovmskb|pmulhuw|psadbw|pshufw|xorps)(?-i)\\b"
+    - statement: "\\b(?i)(movups|movss|movlps|movhlps|movlps|unpcklps|unpckhps|movhps|movlhps|prefetchnta|prefetch0|prefetch1|prefetch2|nop|movaps|cvtpi2ps|cvtsi2ss|cvtps2pi|cvttss2si|cvtps2pi|cvtss2si|ucomiss|comiss|sqrtps|sqrtss|rsqrtps|rsqrtss|rcpps|andps|orps|xorps|addps|addss|mulps|mulss|subps|subss|minps|minss|divps|divss|maxps|maxss|pshufw|ldmxcsr|stmxcsr|sfence|cmpps|cmpss|pinsrw|pextrw|shufps|pmovmskb|pminub|pmaxub|pavgb|pavgw|pmulhuw|movntq|pminsw|pmaxsw|psadbw|maskmovq)(?-i)\\b"
+    - statement: "\\b(?i)(addpd|addsd|addnpd|cmppd|cmpsd)(?-i)\\b"
+    - statement: "\\b(?i)(addpd|addsd|andnpd|andpd|cmppd|cmpsd|comisd|cvtdq2pd|cvtdq2ps|cvtpd2dq|cvtpd2pi|cvtpd2ps|cvtpi2pd|cvtps2dq|cvtps2pd|cvtsd2si|cvtsd2ss|cvtsi2sd|cvtss2sd|cvttpd2dq|cvttpd2pi|cvttps2dq|cvttsd2si|divpd|divsd|maxpd|maxsd|minpd|minsd|movapd|movhpd|movlpd|movmskpd|movsd|movupd|mulpd|mulsd|orpd|shufpd|sqrtpd|sqrtsd|subpd|subsd|ucomisd|unpckhpd|unpcklpd|xorpd)(?-i)\\b"
+    - statement: "\\b(?i)(movdq2q|movdqa|movdqu|movq2dq|paddq|psubq|pmuludq|pshufhw|pshuflw|pshufd|pslldq|psrldq|punpckhqdq|punpcklqdq)(?-i)\\b"
+    - statement: "\\b(?i)(addsubpd|addsubps|haddpd|haddps|hsubpd|hsubps|movddup|movshdup|movsldu)(?-i)\\b"
+    - statement: "\\b(?i)(lddqu)(?-i)\\b"
+    - statement: "\\b(?i)(psignw|psignd|psignb|pshufb|pmulhrsw|pmaddubsw|phsubw|phsubsw|phsubd|phaddw|phaddsw|phaddd|palignr|pabsw|pabsd|pabsb)(?-i)\\b"
+    - statement: "\\b(?i)(dpps|dppd|blendps|blendpd|blendvps|blendvpd|roundps|roundss|roundpd|roundsd|insertps|extractps)(?-i)\\b"
+    - statement: "\\b(?i)(mpsadbw|phminposuw|pmulld|pmuldq|pblendvb|pblendw|pminsb|pmaxsb|pminuw|pmaxuw|pminud|pmaxud|pminsd|pmaxsd|pinsrb|pinsrd/pinsrq|pextrb|pextrw|pextrd/pextrq|pmovsxbw|pmovzxbw|pmovsxbd|pmovzxbd|pmovsxbq|pmovzxbq|pmovsxwd|pmovzxwd|pmovsxwq|pmovzxwq|pmovsxdq|pmovzxdq|ptest|pcmpeqq|packusdw|movntdqa)(?-i)\\b"
+    - statement: "\\b(?i)(extrq|insertq|movntsd|movntss)(?-i)\\b"
+    - statement: "\\b(?i)(crc32|pcmpestri|pcmpestrm|pcmpistri|pcmpistrm|pcmpgtq)(?-i)\\b"
+    - statement: "\\b(?i)(vfmaddpd|vfmaddps|vfmaddsd|vfmaddss|vfmaddsubpd|vfmaddsubps|vfmsubaddpd|vfmsubaddps|vfmsubpd|vfmsubps|vfmsubsd|vfmsubss|vfnmaddpd|vfnmaddps|vfnmaddsd|vfnmaddss|vfnmsubps|vfnmsubsd|vfnmsubss)(?-i)\\b"
+
+      # Crypto
+    - statement: "\\b(?i)(aesenc|aesenclast|aesdec|aesdeclast|aeskeygenassist|aesimc)(?-i)\\b"
+    - statement: "\\b(?i)(sha1rnds4|sha1nexte|sha1msg1|sha1msg2|sha256rnds2|sha256msg1|sha256msg2)(?-i)\\b"
+
+      # Undocumented
+    - statement: "\\b(?i)(aam|aad|salc|icebp|loadall|loadalld|ud1)(?-i)\\b"
+
+      ## Registers
+    - identifier: "\\b(?i)(al|ah|bl|bh|cl|ch|dl|dh|bpl|sil|r8b|r9b|r10b|r11b|dil|spl|r12b|r13b|r14b|r15)(?-i)\\b"
+    - identifier: "\\b(?i)(cw|sw|tw|fp_ds|fp_opc|fp_ip|fp_dp|fp_cs|cs|ss|ds|es|fs|gs|gdtr|idtr|tr|ldtr|ax|bx|cx|dx|bp|si|r8w|r9w|r10w|r11w|di|sp|r12w|r13w|r14w|r15w|ip)(?-i)\\b"
+    - identifier: "\\b(?i)(fp_dp|fp_ip|eax|ebx|ecx|edx|ebp|esi|r8d|r9d|r10d|r11d|edi|esp|r12d|r13d|r14d|r15d|eip|eflags|mxcsr)(?-i)\\b"
+    - identifier: "\\b(?i)(mm0|mm1|mm2|mm3|mm4|mm5|mm6|mm7|rax|rbx|rcx|rdx|rbp|rsi|r8|r9|r10|r11|rdi|rsp|r12|r13|r14|r15|rip|rflags|cr0|cr1|cr2|cr3|cr4|cr5|cr6|cr7|cr8|cr9|cr10|cr11|cr12|cr13|cr14|cr15|msw|dr0|dr1|dr2|dr3|r4|dr5|dr6|dr7|dr8|dr9|dr10|dr11|dr12|dr13|dr14|dr15)(?-i)\\b"
+    - identifier: "\\b(?i)(st0|st1|st2|st3|st4|st5|st6|st7)(?-i)\\b"
+    - identifier: "\\b(?i)(xmm0|xmm1|xmm2|xmm3|xmm4|xmm5|xmm6|xmm7|xmm8|xmm9|xmm10|xmm11|xmm12|xmm13|xmm14|xmm15)(?-i)\\b"
+    - identifier: "\\b(?i)(ymm0|ymm1|ymm2|ymm3|ymm4|ymm5|ymm6|ymm7|ymm8|ymm9|ymm10|ymm11|ymm12|ymm13|ymm14|ymm15)(?-i)\\b"
+    - identifier: "\\b(?i)(zmm0|zmm1|zmm2|zmm3|zmm4|zmm5|zmm6|zmm7|zmm8|zmm9|zmm10|zmm11|zmm12|zmm13|zmm14|zmm15|zmm16|zmm17|zmm18|zmm19|zmm20|zmm21|zmm22|zmm23|zmm24|zmm25|zmm26|zmm27|zmm28|zmm29|zmm30|zmm31)(?-i)\\b"
+
+      ## Constants
+      # Number - it works
+    - constant.number: "\\b(|h|A|0x)+[0-9]+(|h|A)+\\b"
+    - constant.number: "\\b0x[0-9 a-f A-F]+\\b"
+
+      ## Preprocessor (NASM)
+    - preproc: "%+(\\+|\\?|\\?\\?|)[a-z A-Z 0-9]+"
+    - preproc: "%\\[[. a-z A-Z 0-9]*\\]"
+
+      ## Other
+    - statement: "\\b(?i)(extern|global|section|segment|_start|\\.text|\\.data|\\.bss)(?-i)\\b"
+    - statement: "\\b(?i)(db|dw|dd|dq|dt|ddq|do)(?-i)\\b"
+    - identifier: "[a-z A-Z 0-9 _]+:"
+
+    - constant.string:
+        start: "\""
+        end: "\""
+        rules:
+            - constant.specialChar: "\\\\."
+
+    - constant.string:
+        start: "'"
+        end: "'"
+        rules:
+            - constant.specialChar: "\\\\."
+
+    - comment:
+        start: ";"
+        end: "$"
+        rules:
+            - todo: "(TODO|XXX|FIXME):?"

runtime/syntax/c.yaml

@@ -39,9 +39,11 @@ rules:
     - comment:
         start: "//"
         end: "$"
-        rules: []
+        rules:
+            - todo: "(TODO|XXX|FIXME):?"
 
     - comment:
         start: "/\\*"
         end: "\\*/"
-        rules: []
+        rules:
+            - todo: "(TODO|XXX|FIXME):?"

runtime/syntax/caddyfile.yaml

@@ -0,0 +1,21 @@
+filetype: caddyfile
+
+detect:
+    filename: "Caddyfile"
+
+rules:
+    - identifier: "^\\s*\\S+(\\s|$)"
+    - type: "^([\\w.:/-]+,? ?)+[,{]$"
+    - constant.specialChar: "\\s{$"
+    - constant.specialChar: "^\\s*}$"
+    - constant.string:
+        start: "\""
+        end: "\""
+        rules:
+            - constant.specialChar: "\\\\."
+
+    - preproc: "\\{(\\w+|\\$\\w+|%\\w+%)\\}"
+    - comment:
+        start: "#"
+        end: "$"
+        rules: []

runtime/syntax/coffeescript.yaml

@@ -0,0 +1,27 @@
+filetype: coffeescript
+
+detect:
+    filename: "\\.coffee$"
+
+rules:
+    - statement: "[!&|=/*+-<>]|\\b(and|or|is|isnt|not)\\b"
+    - identifier: "([A-Za-z_][A-Za-z0-9_]*:[[:space:]]*(->|\\()|->)"
+    - statement: "[()]"
+    - statement:  "\\b(for|of|continue|break|isnt|null|unless|this|else|if|return)\\b"
+    - statement:  "\\b(try|catch|finally|throw|new|delete|typeof|in|instanceof)\\b"
+    - statement:  "\\b(debugger|switch|while|do|class|extends|super)\\b"
+    - statement:  "\\b(undefined|then|unless|until|loop|of|by|when)\\b"
+    - constant:  "\\b(true|false|yes|no|on|off)\\b"
+    - preproc: "@[A-Za-z0-9_]*"
+
+    - constant.string:
+        start: "\""
+        end: "\""
+        rules:
+            - constant.specialChar: "\\\\."
+
+    - comment:
+        start: "#"
+        end: "$"
+        rules:
+            - todo: "(TODO|XXX|FIXME):?"

runtime/syntax/dockerfile.yaml

@@ -0,0 +1,33 @@
+filetype: dockerfile
+
+detect:
+    filename: "(Dockerfile[^/]*$|\\.dockerfile$)"
+
+rules:
+    ## Keywords
+    - keyword: "(?i)^(FROM|MAINTAINER|RUN|CMD|LABEL|EXPOSE|ENV|ADD|COPY|ENTRYPOINT|VOLUME|USER|WORKDIR|ONBUILD|ARG|HEALTHCHECK|STOPSIGNAL|SHELL)[[:space:]]"
+
+      ## Brackets & parenthesis
+    - statement: "(\\(|\\)|\\[|\\])"
+
+      ## Double ampersand
+    - special: "&&"
+
+      ## Comments
+    - comment:
+        start: "#"
+        end: "$"
+        rules:
+            - todo: "(TODO|XXX|FIXME):?"
+
+    - constant.string:
+        start: "\""
+        end: "\""
+        rules:
+            - constant.specialChar: "\\\\."
+
+    - constant.string:
+        start: "'"
+        end: "'"
+        rules:
+            - constant.specialChar: "\\\\."

runtime/syntax/go.yaml

@@ -1,7 +1,7 @@
 filetype: go
 
 detect:
-  filename: "\\.go$"
+    filename: "\\.go$"
 
 rules:
     - statement: "\\b(break|case|continue|default|else|for|go|goto|if|range|return|switch)\\b"

runtime/syntax/syntax_checker.go

@@ -0,0 +1,30 @@
+package main
+
+import (
+	"fmt"
+	"io/ioutil"
+	"strings"
+
+	"github.com/zyedidia/micro/cmd/micro/highlight"
+)
+
+func main() {
+	files, _ := ioutil.ReadDir(".")
+
+	hadErr := false
+	for _, f := range files {
+		if strings.HasSuffix(f.Name(), ".yaml") {
+			input, _ := ioutil.ReadFile("/Users/zachary/gocode/src/github.com/zyedidia/highlight/syntax_files/" + f.Name())
+			_, err := highlight.ParseDef(input)
+			if err != nil {
+				hadErr = true
+				fmt.Printf("%s:\n", f.Name())
+				fmt.Println(err)
+				continue
+			}
+		}
+	}
+	if !hadErr {
+		fmt.Println("No issues!")
+	}
+}