mirror of
https://github.com/mattermost/mattermost.git
synced 2026-09-21 14:20:41 +08:00
* [MM-24529] Preserve hyphenated compound words in Postgres search Postgres full-text search stripped every hyphen to a space before building a to_tsquery, so an unquoted search for "t-shirt" degraded into a loose "t AND shirt" match instead of the compound-word match the quoted version already got. Fix is Postgres-only; MySQL is gone and Elasticsearch/Bleve use a different tokenizer/query mechanism (verified empirically against a live ES instance - out of scope here). - Remove "-" from the shared specialSearchChars strip list in store.go, and add neutralizeNonWordHyphens (utils.go), which keeps a hyphen only when flanked by a letter/digit on both sides, neutralizing malformed usage (leading/trailing/standalone/repeated) that would otherwise reach to_tsquery as a parse error. - Wire the new helper into post_store.go's tsquery-building path. - file_info_store.go keeps the old hyphen-to-space behavior: real filenames glue a hyphenated name directly to an extension (e.g. "photo-2024.jpg"), which Postgres's parser tokenizes as an opaque "host"-type token that a hyphen-preserving compound query can never match - verified against Postgres. Re-skip the two FileInfo dash acceptance tests with an accurate reason instead of "Not working". - Re-enable the two previously-skipped Postgres post-search dash acceptance tests, and extend them with excluded-term, wildcard, and letters-digits-compound cases. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> * Revert file_info_layer.go changes Restore the two FileInfo dash acceptance tests to their original Skip:true/SkipMessage state, matching master exactly. The underlying fix (file_info_store.go keeps the old hyphen-to-space behavior since Postgres tokenizes hyphenated-filename+extension as an opaque token) is unaffected - this only reverts the test scaffolding tweak. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
253 lines
7.0 KiB
Go
253 lines
7.0 KiB
Go
// Copyright (c) 2015-present Mattermost, Inc. All Rights Reserved.
|
|
// See LICENSE.txt for license information.
|
|
|
|
package sqlstore
|
|
|
|
import (
|
|
"database/sql"
|
|
"errors"
|
|
"fmt"
|
|
"io"
|
|
"maps"
|
|
"net/url"
|
|
"strconv"
|
|
"strings"
|
|
"unicode"
|
|
|
|
"github.com/wiggin77/merror"
|
|
|
|
"github.com/mattermost/mattermost/server/public/model"
|
|
"github.com/mattermost/mattermost/server/public/shared/mlog"
|
|
)
|
|
|
|
// defaultMaxInsertParams is a conservative threshold (76% of PostgreSQL's
|
|
// 65,535 parameter limit) used to chunk bulk INSERT statements so they never
|
|
// overflow the wire-protocol's 16-bit parameter counter.
|
|
const defaultMaxInsertParams = 50_000
|
|
|
|
// chunkSlice splits items into sub-slices sized so that each chunk uses at most
|
|
// maxParams query parameters (columnsPerRow params per item). When the input
|
|
// already fits in one chunk the original slice is returned with zero allocation
|
|
// overhead.
|
|
func chunkSlice[T any](items []T, columnsPerRow int, maxParams int) [][]T {
|
|
if columnsPerRow <= 0 {
|
|
panic(fmt.Sprintf("chunkSlice: columnsPerRow must be > 0, got %d", columnsPerRow))
|
|
}
|
|
if maxParams <= 0 {
|
|
panic(fmt.Sprintf("chunkSlice: maxParams must be > 0, got %d", maxParams))
|
|
}
|
|
if columnsPerRow > maxParams {
|
|
panic(fmt.Sprintf("chunkSlice: columnsPerRow (%d) must be <= maxParams (%d)", columnsPerRow, maxParams))
|
|
}
|
|
if len(items) == 0 {
|
|
return nil
|
|
}
|
|
chunkSize := maxParams / columnsPerRow
|
|
if len(items) <= chunkSize {
|
|
return [][]T{items}
|
|
}
|
|
var chunks [][]T
|
|
for i := 0; i < len(items); i += chunkSize {
|
|
end := min(i+chunkSize, len(items))
|
|
chunks = append(chunks, items[i:end])
|
|
}
|
|
return chunks
|
|
}
|
|
|
|
var escapeLikeSearchChar = []string{
|
|
"%",
|
|
"_",
|
|
}
|
|
|
|
func sanitizeSearchTerm(term string, escapeChar string) string {
|
|
term = strings.Replace(term, escapeChar, "", -1)
|
|
|
|
for _, c := range escapeLikeSearchChar {
|
|
term = strings.Replace(term, c, escapeChar+c, -1)
|
|
}
|
|
|
|
return term
|
|
}
|
|
|
|
// Converts a list of strings into a list of query parameters and a named parameter map that can
|
|
// be used as part of a SQL query.
|
|
func MapStringsToQueryParams(list []string, paramPrefix string) (string, map[string]any) {
|
|
var keys strings.Builder
|
|
params := make(map[string]any, len(list))
|
|
for i, entry := range list {
|
|
if keys.Len() > 0 {
|
|
keys.WriteString(",")
|
|
}
|
|
|
|
key := paramPrefix + strconv.Itoa(i)
|
|
keys.WriteString(":" + key)
|
|
params[key] = entry
|
|
}
|
|
|
|
return "(" + keys.String() + ")", params
|
|
}
|
|
|
|
// finalizeTransactionX ensures a transaction is closed after use, rolling back if not already committed.
|
|
func finalizeTransactionX(transaction *sqlxTxWrapper, perr *error) {
|
|
// Rollback returns sql.ErrTxDone if the transaction was already closed.
|
|
if err := transaction.Rollback(); err != nil && !errors.Is(err, sql.ErrTxDone) {
|
|
*perr = merror.Append(*perr, err)
|
|
}
|
|
}
|
|
|
|
func deferClose(c io.Closer, perr *error) {
|
|
err := c.Close()
|
|
*perr = merror.Append(*perr, err)
|
|
}
|
|
|
|
// removeNonAlphaNumericUnquotedTerms removes all unquoted words that only contain
|
|
// non-alphanumeric chars from given line
|
|
func removeNonAlphaNumericUnquotedTerms(line, separator string) string {
|
|
words := strings.Split(line, separator)
|
|
filteredResult := make([]string, 0, len(words))
|
|
|
|
for _, w := range words {
|
|
if isQuotedWord(w) || containsAlphaNumericChar(w) {
|
|
filteredResult = append(filteredResult, strings.TrimSpace(w))
|
|
}
|
|
}
|
|
return strings.Join(filteredResult, separator)
|
|
}
|
|
|
|
// containsAlphaNumericChar returns true in case any letter or digit is present, false otherwise
|
|
func containsAlphaNumericChar(s string) bool {
|
|
for _, r := range s {
|
|
if unicode.IsLetter(r) || unicode.IsDigit(r) {
|
|
return true
|
|
}
|
|
}
|
|
return false
|
|
}
|
|
|
|
// isQuotedWord return true if the input string is quoted, false otherwise. Ex :-
|
|
//
|
|
// "quoted string" - will return true
|
|
// unquoted string - will return false
|
|
func isQuotedWord(s string) bool {
|
|
if len(s) < 2 {
|
|
return false
|
|
}
|
|
|
|
return s[0] == '"' && s[len(s)-1] == '"'
|
|
}
|
|
|
|
func constructArrayArgs(ids []string) (string, []any) {
|
|
var placeholder strings.Builder
|
|
values := make([]any, 0, len(ids))
|
|
for _, entry := range ids {
|
|
if placeholder.Len() > 0 {
|
|
placeholder.WriteString(",")
|
|
}
|
|
|
|
placeholder.WriteString("?")
|
|
values = append(values, entry)
|
|
}
|
|
|
|
return "(" + placeholder.String() + ")", values
|
|
}
|
|
|
|
func wrapBinaryParamStringMap(ok bool, props model.StringMap) model.StringMap {
|
|
if props == nil {
|
|
props = make(model.StringMap)
|
|
}
|
|
props[model.BinaryParamKey] = strconv.FormatBool(ok)
|
|
return props
|
|
}
|
|
|
|
// morphWriter is a target to pass to the logger instance of morph.
|
|
// For now, everything is just logged at a debug level. If we need to log
|
|
// errors/warnings from the library also, that needs to be seen later.
|
|
type morphWriter struct{}
|
|
|
|
func (l *morphWriter) Write(in []byte) (int, error) {
|
|
mlog.Debug(strings.TrimSpace(string(in)))
|
|
return len(in), nil
|
|
}
|
|
|
|
func DSNHasBinaryParam(dsn string) (bool, error) {
|
|
url, err := url.Parse(dsn)
|
|
if err != nil {
|
|
return false, err
|
|
}
|
|
return url.Query().Get("binary_parameters") == "yes", nil
|
|
}
|
|
|
|
// AppendBinaryFlag updates the byte slice to work using binary_parameters=yes.
|
|
func AppendBinaryFlag(buf []byte) []byte {
|
|
return append([]byte{0x01}, buf...)
|
|
}
|
|
|
|
const maxTokenSize = 50
|
|
|
|
// trimInput limits the string to a max size to prevent clogging up disk space
|
|
// while logging
|
|
func trimInput(input string) string {
|
|
if len(input) > maxTokenSize {
|
|
input = input[:maxTokenSize] + "..."
|
|
}
|
|
return input
|
|
}
|
|
|
|
// rowScanner is the minimal interface needed to iterate over SQL result rows.
|
|
type rowScanner interface {
|
|
Next() bool
|
|
Scan(dest ...any) error
|
|
Err() error
|
|
}
|
|
|
|
// neutralizeNonWordHyphens replaces any '-' that isn't flanked by a word
|
|
// rune (letter or digit) on both sides with a space, so malformed hyphen
|
|
// usage (leading/trailing/standalone/repeated) can't reach to_tsquery, while
|
|
// compound words like "t-shirt" are preserved.
|
|
func neutralizeNonWordHyphens(s string) string {
|
|
if !strings.ContainsRune(s, '-') {
|
|
return s
|
|
}
|
|
runes := []rune(s)
|
|
for i, r := range runes {
|
|
if r != '-' {
|
|
continue
|
|
}
|
|
hasLeft := i > 0 && isWordRune(runes[i-1])
|
|
hasRight := i < len(runes)-1 && isWordRune(runes[i+1])
|
|
if !hasLeft || !hasRight {
|
|
runes[i] = ' '
|
|
}
|
|
}
|
|
return string(runes)
|
|
}
|
|
|
|
// isWordRune reports whether r can be part of a word for hyphen-flanking
|
|
// purposes. Combining marks (e.g. a decomposed accent) count too, since they
|
|
// attach to the preceding base letter rather than acting as a boundary.
|
|
func isWordRune(r rune) bool {
|
|
return unicode.IsLetter(r) || unicode.IsDigit(r) || unicode.IsMark(r)
|
|
}
|
|
|
|
// scanRowsIntoMap scans SQL rows into a map, using a provided scanner function to extract key-value pairs
|
|
func scanRowsIntoMap[K comparable, V any](rows rowScanner, scanner func(rows rowScanner) (K, V, error), defaults map[K]V) (map[K]V, error) {
|
|
results := make(map[K]V, len(defaults))
|
|
|
|
// Initialize with default values if provided
|
|
maps.Copy(results, defaults)
|
|
|
|
for rows.Next() {
|
|
key, value, err := scanner(rows)
|
|
if err != nil {
|
|
return nil, err
|
|
}
|
|
results[key] = value
|
|
}
|
|
|
|
if err := rows.Err(); err != nil {
|
|
return nil, fmt.Errorf("error while iterating rows: %w", err)
|
|
}
|
|
|
|
return results, nil
|
|
}
|