exp/locale/collate: changed API to allow access to different locales through New(),

instead of variables. Several reasons:
- Encourage users of the API to minimize the number of creations and reuse Collate objects.
- Don't rule out the possibility of using initialization code for collators. For some locales
  it will be possible to have very compact representations that can be quickly expanded
  into a proper table on demand.
Other changes:
- Change name of root* vars to main*, as the tables are shared between locales.
- Added Locales() method to get a list of supported locales.

R=r
CC=golang-dev
https://golang.org/cl/6498107

diff --git a/src/pkg/exp/locale/collate/build/builder.go b/src/pkg/exp/locale/collate/build/builder.go
index 401ef24e38795af8463f96347d04d5d5277cef06..97d5e81ddd04a62530c240d6073f930ae9123133 100644 (file)
--- a/src/pkg/exp/locale/collate/build/builder.go
+++ b/src/pkg/exp/locale/collate/build/builder.go
@@ -263,14 +263,31 @@ func (t *Tailoring) Build() (*collate.Collator, error) {
 
 // Print prints the tables for b and all its Tailorings as a Go file
 // that can be included in the Collate package.
-func (b *Builder) Print(w io.Writer) (int, error) {
+func (b *Builder) Print(w io.Writer) (n int, err error) {
+       p := func(nn int, e error) {
+               n += nn
+               if err == nil {
+                       err = e
+               }
+       }
        t, err := b.build()
        if err != nil {
                return 0, err
        }
-       // TODO: support multiple locales
-       n, _, err := t.fprint(w, "root")
-       return n, err
+       p(fmt.Fprintf(w, "var availableLocales = []string{"))
+       for _, loc := range b.locale {
+               p(fmt.Fprintf(w, "%q, ", loc.id))
+       }
+       p(fmt.Fprintln(w, "}\n"))
+       p(fmt.Fprintln(w, "var locales = map[string]tableIndex{"))
+       for _, loc := range b.locale {
+               p(fmt.Fprintf(w, "\t%q: ", loc.id))
+               p(t.fprintIndex(w, loc.index.handle))
+               p(fmt.Fprintln(w, ","))
+       }
+       p(fmt.Fprint(w, "}\n\n"))
+       n, _, err = t.fprint(w, "main")
+       return
 }
 
 // reproducibleFromNFKD checks whether the given expansion could be generated

diff --git a/src/pkg/exp/locale/collate/build/table.go b/src/pkg/exp/locale/collate/build/table.go
index a7973f5510e22bfbd396954c486cdd14f340aa5d..0a290cd00b534aab32d27a695b4e162233d89cfe 100644 (file)
--- a/src/pkg/exp/locale/collate/build/table.go
+++ b/src/pkg/exp/locale/collate/build/table.go
@@ -96,6 +96,21 @@ func (t *table) fprint(w io.Writer, name string) (n, size int, err error) {
        return
 }
 
+func (t *table) fprintIndex(w io.Writer, h *trieHandle) (n int, err error) {
+       p := func(f string, a ...interface{}) {
+               nn, e := fmt.Fprintf(w, f, a...)
+               n += nn
+               if err == nil {
+                       err = e
+               }
+       }
+       p("tableIndex{\n")
+       p("\t\tlookupOffset: 0x%x,\n", h.lookupStart)
+       p("\t\tvaluesOffset: 0x%x,\n", h.valueStart)
+       p("\t}")
+       return
+}
+
 func printColElems(w io.Writer, a []uint32, name string) (n, sz int, err error) {
        p := func(f string, a ...interface{}) {
                nn, e := fmt.Fprintf(w, f, a...)

diff --git a/src/pkg/exp/locale/collate/collate.go b/src/pkg/exp/locale/collate/collate.go
index 081c3b4ce733ab7b6f4a5b1915ca323c9beab239..5853b710c337f025980c6eee8d2fe98e5d040a9e 100644 (file)
--- a/src/pkg/exp/locale/collate/collate.go
+++ b/src/pkg/exp/locale/collate/collate.go
@@ -85,6 +85,27 @@ type Collator struct {
        t *table
 }
 
+// Locales returns the list of locales for which collating differs from its parent locale.
+func Locales() []string {
+       return availableLocales
+}
+
+// New returns a new Collator initialized for the given locale.
+func New(loc string) *Collator {
+       // TODO: handle locale selection according to spec.
+       t := &mainTable
+       if loc != "" {
+               if idx, ok := locales[loc]; ok {
+                       t = mainTable.indexedTable(idx)
+               }
+       }
+       return &Collator{
+               Strength: Quaternary,
+               f:        norm.NFD,
+               t:        t,
+       }
+}
+
 // SetVariableTop sets all runes with primary strength less than the primary
 // strength of r to be variable and thus affected by alternate handling.
 func (c *Collator) SetVariableTop(r rune) {

diff --git a/src/pkg/exp/locale/collate/maketables.go b/src/pkg/exp/locale/collate/maketables.go
index 481c64c457f6de3e2fa2ff6d7c25cb55073c30b0..7e05ec87a57b02f707cbc3349b69c941d08ce5d7 100644 (file)
--- a/src/pkg/exp/locale/collate/maketables.go
+++ b/src/pkg/exp/locale/collate/maketables.go
@@ -658,7 +658,7 @@ func insertCollation(builder *build.Builder, locale string, c *Collation) {
 }
 
 func testCollator(c *collate.Collator) {
-       c0 := collate.Root
+       c0 := collate.New("")
 
        // iterator over all characters for all locales and check
        // whether Key is equal.
@@ -682,19 +682,6 @@ func testCollator(c *collate.Collator) {
        fmt.Println("PASS")
 }
 
-// TODO: move this functionality to exp/locale/collate/build.
-func printCollators(c *collate.Collator) {
-       const name = "Root"
-       fmt.Printf("var _%s = Collator{\n", name)
-       fmt.Printf("\tStrength: %v,\n", c.Strength)
-       fmt.Printf("\tf: norm.NFD,\n")
-       fmt.Printf("\tt: &%sTable,\n", strings.ToLower(name))
-       fmt.Printf("}\n\n")
-       fmt.Printf("var (\n")
-       fmt.Printf("\t%s = _%s\n", name, name)
-       fmt.Printf(")\n\n")
-}
-
 func main() {
        flag.Parse()
        b := build.NewBuilder()
@@ -725,9 +712,6 @@ func main() {
                fmt.Printf("package %s\n", *pkg)
                if tables.contains("collate") {
                        fmt.Println("")
-                       fmt.Println(`import "exp/norm"`)
-                       fmt.Println("")
-                       printCollators(c)
                        _, err = b.Print(os.Stdout)
                        failOnError(err)
                }

diff --git a/src/pkg/exp/locale/collate/table.go b/src/pkg/exp/locale/collate/table.go
index b2a5b6231642c0f404a1cb30e5a57e1c85381e02..c25799b98b231d5b13e78abf969fd445bdabaf93 100644 (file)
--- a/src/pkg/exp/locale/collate/table.go
+++ b/src/pkg/exp/locale/collate/table.go
@@ -9,6 +9,13 @@ import (
        "unicode/utf8"
 )
 
+// tableIndex holds information for constructing a table
+// for a certain locale based on the main table.
+type tableIndex struct {
+       lookupOffset uint32
+       valuesOffset uint32
+}
+
 // table holds all collation data for a given collation ordering.
 type table struct {
        index trie // main trie
@@ -23,6 +30,13 @@ type table struct {
        variableTop    uint32
 }
 
+func (t *table) indexedTable(idx tableIndex) *table {
+       nt := *t
+       nt.index.index0 = t.index.index[idx.lookupOffset*blockSize:]
+       nt.index.values0 = t.index.values[idx.valuesOffset*blockSize:]
+       return &nt
+}
+
 // appendNext appends the weights corresponding to the next rune or 
 // contraction in s.  If a contraction is matched to a discontinuous
 // sequence of runes, the weights for the interstitial runes are 

diff --git a/src/pkg/exp/locale/collate/tables.go b/src/pkg/exp/locale/collate/tables.go
index 30fe862c3d47090a2344557c3bc32c916f96213f..f8d469a56bcc547cc9dbadfa5b2141061913266a 100644 (file)
--- a/src/pkg/exp/locale/collate/tables.go
+++ b/src/pkg/exp/locale/collate/tables.go
@@ -5,29 +5,330 @@
 
 package collate
 
-import "exp/norm"
+var availableLocales = []string{"af", "ar", "as", "az", "be", "bg", "bn", "ca", "cs", "cy", "da", "de", "dz", "el", "en_US_POSIX", "eo", "es", "et", "fa", "fi", "fil", "fo", "fr_CA", "gu", "ha", "haw", "he", "hi", "hr", "hu", "hy", "ig", "is", "ja", "kk", "kl", "km", "kn", "ko", "kok", "ln", "lt", "lv", "mk", "ml", "mr", "mt", "my", "nb", "nn", "nso", "om", "or", "pa", "pl", "ps", "ro", "root", "ru", "se", "si", "sk", "sl", "sq", "sr", "sv", "ta", "te", "th", "tn", "tr", "uk", "ur", "vi", "wae", "yo", "zh"}
 
-var _Root = Collator{
-       Strength: 3,
-       f:        norm.NFD,
-       t:        &rootTable,
+var locales = map[string]tableIndex{
+       "af": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ar": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "as": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "az": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "be": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "bg": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "bn": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ca": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "cs": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "cy": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "da": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "de": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "dz": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "el": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "en_US_POSIX": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "eo": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "es": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "et": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "fa": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "fi": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "fil": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "fo": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "fr_CA": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "gu": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ha": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "haw": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "he": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "hi": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "hr": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "hu": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "hy": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ig": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "is": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ja": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "kk": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "kl": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "km": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "kn": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ko": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "kok": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ln": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "lt": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "lv": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "mk": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ml": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "mr": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "mt": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "my": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "nb": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "nn": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "nso": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "om": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "or": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "pa": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "pl": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ps": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ro": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "root": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ru": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "se": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "si": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "sk": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "sl": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "sq": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "sr": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "sv": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ta": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "te": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "th": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "tn": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "tr": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "uk": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "ur": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "vi": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "wae": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "yo": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
+       "zh": tableIndex{
+               lookupOffset: 0x13,
+               valuesOffset: 0x0,
+       },
 }
 
-var (
-       Root = _Root
-)
-
-var rootTable = table{
-       trie{rootLookup[1216:], rootValues[0:], rootLookup[:], rootValues[:]},
-       rootExpandElem[:],
-       contractTrieSet(rootCTEntries[:]),
-       rootContractElem[:],
+var mainTable = table{
+       trie{mainLookup[1216:], mainValues[0:], mainLookup[:], mainValues[:]},
+       mainExpandElem[:],
+       contractTrieSet(mainCTEntries[:]),
+       mainContractElem[:],
        9,
        0x2ED,
 }
 
-// rootExpandElem: 4642 entries, 18568 bytes
-var rootExpandElem = [4642]uint32{
+// mainExpandElem: 4642 entries, 18568 bytes
+var mainExpandElem = [4642]uint32{
        // Block 0, offset 0x0
        0x00000002, 0x8000A31A, 0x8000AD1A, 0x00000002, 0x8000A51A, 0x8000AD1A,
        0x00000002, 0x8000A718, 0x8000AD18, 0x00000002, 0x8000A71A, 0x8000AD1A,
@@ -901,8 +1202,8 @@ var rootExpandElem = [4642]uint32{
        0x80015F1F, 0x00000002, 0x033F3804, 0x8001601F,
 }
 
-// rootContractElem: 799 entries, 3196 bytes
-var rootContractElem = [799]uint32{
+// mainContractElem: 799 entries, 3196 bytes
+var mainContractElem = [799]uint32{
        // Block 0, offset 0x0
        0x4016C420, 0xE0000789, 0xE0000789, 0x002D8808, 0xE000078F, 0xE000078F,
        0x40194320, 0x40194720, 0x40194B20, 0x00328608, 0x00328E08, 0x00329608,
@@ -1056,9 +1357,9 @@ var rootContractElem = [799]uint32{
        0x40283220,
 }
 
-// rootValues: 25408 entries, 101632 bytes
+// mainValues: 25408 entries, 101632 bytes
 // Block 2 is the null block.
-var rootValues = [25408]uint32{
+var mainValues = [25408]uint32{
        // Block 0x0, offset 0x0
        0x0000: 0x80000000, 0x0001: 0x80000000, 0x0002: 0x80000000, 0x0003: 0x80000000,
        0x0004: 0x80000000, 0x0005: 0x80000000, 0x0006: 0x80000000, 0x0007: 0x80000000,
@@ -7047,9 +7348,9 @@ var rootValues = [25408]uint32{
        0x632c: 0x80000000, 0x632d: 0x80000000, 0x632e: 0x80000000, 0x632f: 0x80000000,
 }
 
-// rootLookup: 1472 entries, 2944 bytes
+// mainLookup: 1472 entries, 2944 bytes
 // Block 0 is the null block.
-var rootLookup = [1472]uint16{
+var mainLookup = [1472]uint16{
        // Block 0x0, offset 0x0
        // Block 0x1, offset 0x40
        // Block 0x2, offset 0x80
@@ -7144,8 +7445,8 @@ var rootLookup = [1472]uint16{
        0x5b0: 0x11, 0x5b3: 0x13,
 }
 
-// rootCTEntries: 126 entries, 504 bytes
-var rootCTEntries = [126]struct{ l, h, n, i uint8 }{
+// mainCTEntries: 126 entries, 504 bytes
+var mainCTEntries = [126]struct{ l, h, n, i uint8 }{
        {0xCE, 0x1, 1, 255},
        {0xC2, 0x0, 1, 255},
        {0xB7, 0xB7, 0, 1},
@@ -7274,4 +7575,4 @@ var rootCTEntries = [126]struct{ l, h, n, i uint8 }{
        {0xB5, 0xB5, 0, 1},
 }
 
-// Total size of rootTable is 126988 bytes
+// Total size of mainTable is 126988 bytes