gosora/query_gen/utils.go

523 lines
12 KiB
Go
Raw Normal View History

/*
*
* Query Generator Library
* WIP Under Construction
Basic search now works for the Search & Filter Widget. ElasticSearch has been temporarily delayed, so I can push through this update. Added the three month time range to the analytics panes. Began work on adding new graphs to the analytics panes. Began work on the ElasticSearch adapter for the search system. Added the currently limited AddKey method to the database adapters. Expanded upon the column parsing logic in the database adapters to ease the use of InsertSelects. Added the BulkGet method to TopicCache. Added the BulkGetMap method to TopicStore. TopicStore methods should now properly retrieve lastReplyBy. Added the panel_analytics_script template to de-dupe part of the analytics logic. We plan to tidy this up further, but for now, it'll suffice. Added plugin_sendmail and plugin_hyperdrive to the continuous integration test list. Tweaked the width and heights of the textareas for the Widget Editor. Added the AddKey method to *qgen.builder Fixed a bug where using the inline forum editor would crash Gosora and wouldn't set the preset permissions for that forum properly. Added DotBot to the user agent analytics. Invisibles should be better handled when they're encountered now in user agent strings. Unknown language ISO Codes in headers now have the requests fully logged for debugging purposes. Shortened some of the pointer receiver names. Shortened some variable names. Added the dotbot phrase. Added the panel_statistics_time_range_three_months phrase. Added gopkg.in/olivere/elastic.v6 as a dependency. You will need to run the patcher or updater for this commit.
2019-02-23 06:29:19 +00:00
* Copyright Azareal 2017 - 2020
*
*/
package qgen
import (
"os"
"strings"
)
Basic search now works for the Search & Filter Widget. ElasticSearch has been temporarily delayed, so I can push through this update. Added the three month time range to the analytics panes. Began work on adding new graphs to the analytics panes. Began work on the ElasticSearch adapter for the search system. Added the currently limited AddKey method to the database adapters. Expanded upon the column parsing logic in the database adapters to ease the use of InsertSelects. Added the BulkGet method to TopicCache. Added the BulkGetMap method to TopicStore. TopicStore methods should now properly retrieve lastReplyBy. Added the panel_analytics_script template to de-dupe part of the analytics logic. We plan to tidy this up further, but for now, it'll suffice. Added plugin_sendmail and plugin_hyperdrive to the continuous integration test list. Tweaked the width and heights of the textareas for the Widget Editor. Added the AddKey method to *qgen.builder Fixed a bug where using the inline forum editor would crash Gosora and wouldn't set the preset permissions for that forum properly. Added DotBot to the user agent analytics. Invisibles should be better handled when they're encountered now in user agent strings. Unknown language ISO Codes in headers now have the requests fully logged for debugging purposes. Shortened some of the pointer receiver names. Shortened some variable names. Added the dotbot phrase. Added the panel_statistics_time_range_three_months phrase. Added gopkg.in/olivere/elastic.v6 as a dependency. You will need to run the patcher or updater for this commit.
2019-02-23 06:29:19 +00:00
// TODO: Add support for numbers and strings?
func processColumns(colStr string) (columns []DBColumn) {
if colStr == "" {
return columns
}
colStr = strings.Replace(colStr, " as ", " AS ", -1)
for _, segment := range strings.Split(colStr, ",") {
var outCol DBColumn
dotHalves := strings.Split(strings.TrimSpace(segment), ".")
var halves []string
if len(dotHalves) == 2 {
outCol.Table = dotHalves[0]
halves = strings.Split(dotHalves[1], " AS ")
} else {
halves = strings.Split(dotHalves[0], " AS ")
}
halves[0] = strings.TrimSpace(halves[0])
if len(halves) == 2 {
outCol.Alias = strings.TrimSpace(halves[1])
}
if halves[0][len(halves[0])-1] == ')' {
outCol.Type = TokenFunc
} else if halves[0] == "?" {
outCol.Type = TokenSub
} else {
outCol.Type = TokenColumn
}
outCol.Left = halves[0]
columns = append(columns, outCol)
}
return columns
}
// TODO: Allow order by statements without a direction
func processOrderby(orderStr string) (order []DBOrder) {
if orderStr == "" {
return order
}
for _, segment := range strings.Split(orderStr, ",") {
var outOrder DBOrder
halves := strings.Split(strings.TrimSpace(segment), " ")
if len(halves) != 2 {
continue
}
outOrder.Column = halves[0]
outOrder.Order = strings.ToLower(halves[1])
order = append(order, outOrder)
}
return order
}
func processJoiner(joinStr string) (joiner []DBJoiner) {
if joinStr == "" {
return joiner
}
joinStr = strings.Replace(joinStr, " on ", " ON ", -1)
joinStr = strings.Replace(joinStr, " and ", " AND ", -1)
for _, segment := range strings.Split(joinStr, " AND ") {
var outJoin DBJoiner
var parseOffset int
var left, right string
left, parseOffset = getIdentifier(segment, parseOffset)
outJoin.Operator, parseOffset = getOperator(segment, parseOffset+1)
right, parseOffset = getIdentifier(segment, parseOffset+1)
leftColumn := strings.Split(left, ".")
rightColumn := strings.Split(right, ".")
outJoin.LeftTable = strings.TrimSpace(leftColumn[0])
outJoin.RightTable = strings.TrimSpace(rightColumn[0])
outJoin.LeftColumn = strings.TrimSpace(leftColumn[1])
outJoin.RightColumn = strings.TrimSpace(rightColumn[1])
joiner = append(joiner, outJoin)
}
return joiner
}
func (wh *DBWhere) parseNumber(seg string, i int) int {
//var buffer string
si := i
l := 0
for ; i < len(seg); i++ {
ch := seg[i]
if '0' <= ch && ch <= '9' {
//buffer += string(ch)
l++
} else {
i--
var str string
if l != 0 {
str = seg[si : si+l]
}
wh.Expr = append(wh.Expr, DBToken{str, TokenNumber})
return i
}
}
return i
}
func (wh *DBWhere) parseColumn(seg string, i int) int {
//var buffer string
si := i
l := 0
for ; i < len(seg); i++ {
ch := seg[i]
2017-11-13 00:36:31 +00:00
switch {
case ('a' <= ch && ch <= 'z') || ('A' <= ch && ch <= 'Z') || ch == '.' || ch == '_':
//buffer += string(ch)
l++
case ch == '(':
var str string
if l != 0 {
str = seg[si : si+l]
}
return wh.parseFunction(seg, str, i)
2017-11-13 00:36:31 +00:00
default:
i--
var str string
if l != 0 {
str = seg[si : si+l]
}
wh.Expr = append(wh.Expr, DBToken{str, TokenColumn})
return i
}
}
return i
}
func (wh *DBWhere) parseFunction(seg string, buffer string, i int) int {
preI := i
i = skipFunctionCall(seg, i-1)
buffer += seg[preI:i] + string(seg[i])
wh.Expr = append(wh.Expr, DBToken{buffer, TokenFunc})
return i
}
func (wh *DBWhere) parseString(seg string, i int) int {
//var buffer string
2017-11-13 00:31:46 +00:00
i++
si := i
l := 0
for ; i < len(seg); i++ {
ch := seg[i]
if ch != '\'' {
//buffer += string(ch)
l++
2017-11-13 00:31:46 +00:00
} else {
var str string
if l != 0 {
str = seg[si : si+l]
}
wh.Expr = append(wh.Expr, DBToken{str, TokenString})
2017-11-13 00:31:46 +00:00
return i
}
}
return i
}
func (wh *DBWhere) parseOperator(seg string, i int) int {
//var buffer string
si := i
l := 0
for ; i < len(seg); i++ {
ch := seg[i]
if isOpByte(ch) {
//buffer += string(ch)
l++
2017-11-13 00:31:46 +00:00
} else {
i--
var str string
if l != 0 {
str = seg[si : si+l]
}
wh.Expr = append(wh.Expr, DBToken{str, TokenOp})
2017-11-13 00:31:46 +00:00
return i
}
}
return i
}
// TODO: Make this case insensitive
func normalizeAnd(in string) string {
Fixed a bug where it would use the wrong templates for Tempra Simple, Tempra Cursive, and Shadow Likes are now done over AJAX. Posts you have liked are now visually differentiated from those which you have not. Added support for OR to the where parser. || and && now get translated to OR and AND in the where parser. Added support for ( and ) in the where parser. Added an adapter and builder method for getting the database version. Multiple wheres can now be chained with the micro and accumulator builders. Added the In method to the accumulator select builder. Added the GetConn method to the builder. /uploads/ files should now get cached properly. Added more tooltips for topic titles and usernames. Fixed a bug in the runners where old stale templates would be served. Fixed a bug where liking topics didn't work. Began moving the database initialisation logic out of {adapter}.go and into querygen. Tweaked the alert direction to show the newest alerts rather than the oldest. Tweaked the WS JS to have it handle messages more efficiently. Partially fixed an issue where inline edited posts would lack newlines until the page is refreshed. Used arrow functions in a few places in global.js to save a few characters. Schema: Added the liked, oldestItemLikedCreatedAt and lastLiked columns to the users table. Added the createdAt column to the likes table. MySQL Update Queries: ALTER TABLE `users` ADD COLUMN `liked` INT NOT NULL DEFAULT '0' AFTER `topics`; ALTER TABLE `users` ADD COLUMN `oldestItemLikedCreatedAt` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP AFTER `liked`; ALTER TABLE `users` ADD COLUMN `lastLiked` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP AFTER `oldestItemLikedCreatedAt`; ALTER TABLE `likes` ADD COLUMN `createdAt` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP AFTER `sentBy`; delete from `likes`; delete from `activity_stream` where `event` = 'like'; delete from `activity_stream_matches` where `asid` not in(select `asid` from `activity_stream`); update `topics` set `likeCount` = 0; update `replies` set `likeCount` = 0;
2018-03-31 05:25:27 +00:00
in = strings.Replace(in, " and ", " AND ", -1)
return strings.Replace(in, " && ", " AND ", -1)
}
func normalizeOr(in string) string {
in = strings.Replace(in, " or ", " OR ", -1)
return strings.Replace(in, " || ", " OR ", -1)
2017-11-13 00:31:46 +00:00
}
// TODO: Write tests for this
func processWhere(whereStr string) (where []DBWhere) {
if whereStr == "" {
return where
}
whereStr = normalizeAnd(whereStr)
whereStr = normalizeOr(whereStr)
for _, seg := range strings.Split(whereStr, " AND ") {
tmpWhere := &DBWhere{[]DBToken{}}
seg += ")"
for i := 0; i < len(seg); i++ {
ch := seg[i]
switch {
case '0' <= ch && ch <= '9':
i = tmpWhere.parseNumber(seg, i)
Fixed a bug where it would use the wrong templates for Tempra Simple, Tempra Cursive, and Shadow Likes are now done over AJAX. Posts you have liked are now visually differentiated from those which you have not. Added support for OR to the where parser. || and && now get translated to OR and AND in the where parser. Added support for ( and ) in the where parser. Added an adapter and builder method for getting the database version. Multiple wheres can now be chained with the micro and accumulator builders. Added the In method to the accumulator select builder. Added the GetConn method to the builder. /uploads/ files should now get cached properly. Added more tooltips for topic titles and usernames. Fixed a bug in the runners where old stale templates would be served. Fixed a bug where liking topics didn't work. Began moving the database initialisation logic out of {adapter}.go and into querygen. Tweaked the alert direction to show the newest alerts rather than the oldest. Tweaked the WS JS to have it handle messages more efficiently. Partially fixed an issue where inline edited posts would lack newlines until the page is refreshed. Used arrow functions in a few places in global.js to save a few characters. Schema: Added the liked, oldestItemLikedCreatedAt and lastLiked columns to the users table. Added the createdAt column to the likes table. MySQL Update Queries: ALTER TABLE `users` ADD COLUMN `liked` INT NOT NULL DEFAULT '0' AFTER `topics`; ALTER TABLE `users` ADD COLUMN `oldestItemLikedCreatedAt` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP AFTER `liked`; ALTER TABLE `users` ADD COLUMN `lastLiked` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP AFTER `oldestItemLikedCreatedAt`; ALTER TABLE `likes` ADD COLUMN `createdAt` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP AFTER `sentBy`; delete from `likes`; delete from `activity_stream` where `event` = 'like'; delete from `activity_stream_matches` where `asid` not in(select `asid` from `activity_stream`); update `topics` set `likeCount` = 0; update `replies` set `likeCount` = 0;
2018-03-31 05:25:27 +00:00
// TODO: Sniff the third byte offset from char or it's non-existent to avoid matching uppercase strings which start with OR
case ch == 'O' && (i+1) < len(seg) && seg[i+1] == 'R':
tmpWhere.Expr = append(tmpWhere.Expr, DBToken{"OR", TokenOr})
Fixed a bug where it would use the wrong templates for Tempra Simple, Tempra Cursive, and Shadow Likes are now done over AJAX. Posts you have liked are now visually differentiated from those which you have not. Added support for OR to the where parser. || and && now get translated to OR and AND in the where parser. Added support for ( and ) in the where parser. Added an adapter and builder method for getting the database version. Multiple wheres can now be chained with the micro and accumulator builders. Added the In method to the accumulator select builder. Added the GetConn method to the builder. /uploads/ files should now get cached properly. Added more tooltips for topic titles and usernames. Fixed a bug in the runners where old stale templates would be served. Fixed a bug where liking topics didn't work. Began moving the database initialisation logic out of {adapter}.go and into querygen. Tweaked the alert direction to show the newest alerts rather than the oldest. Tweaked the WS JS to have it handle messages more efficiently. Partially fixed an issue where inline edited posts would lack newlines until the page is refreshed. Used arrow functions in a few places in global.js to save a few characters. Schema: Added the liked, oldestItemLikedCreatedAt and lastLiked columns to the users table. Added the createdAt column to the likes table. MySQL Update Queries: ALTER TABLE `users` ADD COLUMN `liked` INT NOT NULL DEFAULT '0' AFTER `topics`; ALTER TABLE `users` ADD COLUMN `oldestItemLikedCreatedAt` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP AFTER `liked`; ALTER TABLE `users` ADD COLUMN `lastLiked` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP AFTER `oldestItemLikedCreatedAt`; ALTER TABLE `likes` ADD COLUMN `createdAt` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP AFTER `sentBy`; delete from `likes`; delete from `activity_stream` where `event` = 'like'; delete from `activity_stream_matches` where `asid` not in(select `asid` from `activity_stream`); update `topics` set `likeCount` = 0; update `replies` set `likeCount` = 0;
2018-03-31 05:25:27 +00:00
i += 1
case ch == 'N' && (i+2) < len(seg) && seg[i+1] == 'O' && seg[i+2] == 'T':
tmpWhere.Expr = append(tmpWhere.Expr, DBToken{"NOT", TokenNot})
i += 2
case ch == 'L' && (i+3) < len(seg) && seg[i+1] == 'I' && seg[i+2] == 'K' && seg[i+3] == 'E':
tmpWhere.Expr = append(tmpWhere.Expr, DBToken{"LIKE", TokenLike})
i += 3
case ('a' <= ch && ch <= 'z') || ('A' <= ch && ch <= 'Z') || ch == '_':
i = tmpWhere.parseColumn(seg, i)
case ch == '\'':
i = tmpWhere.parseString(seg, i)
case ch == ')' && i < (len(seg)-1):
tmpWhere.Expr = append(tmpWhere.Expr, DBToken{")", TokenOp})
case isOpByte(ch):
i = tmpWhere.parseOperator(seg, i)
case ch == '?':
tmpWhere.Expr = append(tmpWhere.Expr, DBToken{"?", TokenSub})
}
}
where = append(where, *tmpWhere)
}
return where
}
func (set *DBSetter) parseNumber(seg string, i int) int {
//var buffer string
si := i
l := 0
for ; i < len(seg); i++ {
ch := seg[i]
if '0' <= ch && ch <= '9' {
//buffer += string(ch)
l++
2017-11-13 00:31:46 +00:00
} else {
var str string
if l != 0 {
str = seg[si : si+l]
}
set.Expr = append(set.Expr, DBToken{str, TokenNumber})
2017-11-13 00:31:46 +00:00
return i
}
}
return i
}
func (set *DBSetter) parseColumn(seg string, i int) int {
//var buffer string
si := i
l := 0
for ; i < len(seg); i++ {
ch := seg[i]
switch {
case ('a' <= ch && ch <= 'z') || ('A' <= ch && ch <= 'Z') || ch == '_':
//buffer += string(ch)
l++
case ch == '(':
var str string
if l != 0 {
str = seg[si : si+l]
}
return set.parseFunction(seg, str, i)
default:
2017-11-13 00:31:46 +00:00
i--
var str string
if l != 0 {
str = seg[si : si+l]
}
set.Expr = append(set.Expr, DBToken{str, TokenColumn})
2017-11-13 00:31:46 +00:00
return i
}
}
return i
}
func (set *DBSetter) parseFunction(segment string, buffer string, i int) int {
preI := i
2017-11-13 00:31:46 +00:00
i = skipFunctionCall(segment, i-1)
buffer += segment[preI:i] + string(segment[i])
set.Expr = append(set.Expr, DBToken{buffer, TokenFunc})
2017-11-13 00:31:46 +00:00
return i
}
func (set *DBSetter) parseString(seg string, i int) int {
//var buffer string
2017-11-13 00:31:46 +00:00
i++
si := i
l := 0
for ; i < len(seg); i++ {
ch := seg[i]
if ch != '\'' {
//buffer += string(ch)
l++
2017-11-13 00:31:46 +00:00
} else {
var str string
if l != 0 {
str = seg[si : si+l]
}
set.Expr = append(set.Expr, DBToken{str, TokenString})
2017-11-13 00:31:46 +00:00
return i
}
}
return i
}
func (set *DBSetter) parseOperator(seg string, i int) int {
//var buffer string
si := i
l := 0
for ; i < len(seg); i++ {
ch := seg[i]
if isOpByte(ch) {
//buffer += string(ch)
l++
2017-11-13 00:31:46 +00:00
} else {
i--
var str string
if l != 0 {
str = seg[si : si+l]
}
set.Expr = append(set.Expr, DBToken{str, TokenOp})
2017-11-13 00:31:46 +00:00
return i
}
}
return i
}
func processSet(setstr string) (setter []DBSetter) {
if setstr == "" {
return setter
}
// First pass, splitting the string by commas while ignoring the innards of functions
var setset []string
var buffer string
var lastItem int
setstr += ","
for i := 0; i < len(setstr); i++ {
if setstr[i] == '(' {
i = skipFunctionCall(setstr, i-1)
setset = append(setset, setstr[lastItem:i+1])
buffer = ""
lastItem = i + 2
} else if setstr[i] == ',' && buffer != "" {
setset = append(setset, buffer)
buffer = ""
lastItem = i + 1
} else if (setstr[i] > 32) && setstr[i] != ',' && setstr[i] != ')' {
buffer += string(setstr[i])
}
}
// Second pass. Break this setitem into manageable chunks
for _, setitem := range setset {
halves := strings.Split(setitem, "=")
if len(halves) != 2 {
continue
}
2017-11-13 00:31:46 +00:00
tmpSetter := &DBSetter{Column: strings.TrimSpace(halves[0])}
segment := halves[1] + ")"
for i := 0; i < len(segment); i++ {
ch := segment[i]
switch {
case '0' <= ch && ch <= '9':
2017-11-13 00:31:46 +00:00
i = tmpSetter.parseNumber(segment, i)
case ('a' <= ch && ch <= 'z') || ('A' <= ch && ch <= 'Z') || ch == '_':
2017-11-13 00:31:46 +00:00
i = tmpSetter.parseColumn(segment, i)
case ch == '\'':
2017-11-13 00:31:46 +00:00
i = tmpSetter.parseString(segment, i)
case isOpByte(ch):
2017-11-13 00:31:46 +00:00
i = tmpSetter.parseOperator(segment, i)
case ch == '?':
tmpSetter.Expr = append(tmpSetter.Expr, DBToken{"?", TokenSub})
}
}
2017-11-13 00:31:46 +00:00
setter = append(setter, *tmpSetter)
}
return setter
}
func processLimit(limitStr string) (limit DBLimit) {
halves := strings.Split(limitStr, ",")
if len(halves) == 2 {
limit.Offset = halves[0]
limit.MaxCount = halves[1]
} else {
limit.MaxCount = halves[0]
}
return limit
}
func isOpByte(ch byte) bool {
return ch == '<' || ch == '>' || ch == '=' || ch == '!' || ch == '*' || ch == '%' || ch == '+' || ch == '-' || ch == '/' || ch == '(' || ch == ')'
}
func isOpRune(ch rune) bool {
return ch == '<' || ch == '>' || ch == '=' || ch == '!' || ch == '*' || ch == '%' || ch == '+' || ch == '-' || ch == '/' || ch == '(' || ch == ')'
}
func processFields(fieldStr string) (fields []DBField) {
if fieldStr == "" {
return fields
}
var buffer string
var lastItem int
fieldStr += ","
for i := 0; i < len(fieldStr); i++ {
ch := fieldStr[i]
if ch == '(' {
i = skipFunctionCall(fieldStr, i-1)
fields = append(fields, DBField{Name: fieldStr[lastItem : i+1], Type: getIdentifierType(fieldStr[lastItem : i+1])})
buffer = ""
lastItem = i + 2
} else if ch == ',' && buffer != "" {
fields = append(fields, DBField{Name: buffer, Type: getIdentifierType(buffer)})
buffer = ""
lastItem = i + 1
} else if (ch >= 32) && ch != ',' && ch != ')' {
buffer += string(ch)
}
}
return fields
}
func getIdentifierType(iden string) int {
if ('a' <= iden[0] && iden[0] <= 'z') || ('A' <= iden[0] && iden[0] <= 'Z') {
if iden[len(iden)-1] == ')' {
return IdenFunc
}
return IdenColumn
}
if iden[0] == '\'' || iden[0] == '"' {
return IdenString
}
return IdenLiteral
}
func getIdentifier(seg string, startOffset int) (out string, i int) {
seg = strings.TrimSpace(seg)
seg += " " // Avoid overflow bugs with slicing
for i = startOffset; i < len(seg); i++ {
ch := seg[i]
if ch == '(' {
i = skipFunctionCall(seg, i)
return strings.TrimSpace(seg[startOffset:i]), (i - 1)
}
if (ch == ' ' || isOpByte(ch)) && i != startOffset {
return strings.TrimSpace(seg[startOffset:i]), (i - 1)
}
}
return strings.TrimSpace(seg[startOffset:]), (i - 1)
}
func getOperator(seg string, startOffset int) (out string, i int) {
seg = strings.TrimSpace(seg)
seg += " " // Avoid overflow bugs with slicing
for i = startOffset; i < len(seg); i++ {
if !isOpByte(seg[i]) && i != startOffset {
return strings.TrimSpace(seg[startOffset:i]), (i - 1)
}
}
return strings.TrimSpace(seg[startOffset:]), (i - 1)
}
func skipFunctionCall(data string, index int) int {
var braceCount int
for ; index < len(data); index++ {
char := data[index]
if char == '(' {
braceCount++
} else if char == ')' {
braceCount--
if braceCount == 0 {
return index
}
}
}
return index
}
func writeFile(name, content string) (err error) {
f, err := os.Create(name)
if err != nil {
return err
}
_, err = f.WriteString(content)
if err != nil {
return err
}
err = f.Sync()
if err != nil {
return err
}
return f.Close()
}