Implement PCRE Matching (prefer left-branch) #2
@@ -2,7 +2,6 @@ package regex
|
|||||||
|
|
||||||
import (
|
import (
|
||||||
"fmt"
|
"fmt"
|
||||||
"sort"
|
|
||||||
)
|
)
|
||||||
|
|
||||||
// A Match represents a match found by the regex in a given string.
|
// A Match represents a match found by the regex in a given string.
|
||||||
@@ -69,30 +68,6 @@ func getZeroGroup(m Match) Group {
|
|||||||
return m[0]
|
return m[0]
|
||||||
}
|
}
|
||||||
|
|
||||||
// Prunes the slice by removing overlapping indices.
|
|
||||||
func pruneIndices(indices []Match) []Match {
|
|
||||||
// First, sort the slice by the start indices
|
|
||||||
sort.Slice(indices, func(i, j int) bool {
|
|
||||||
return indices[i][0].StartIdx < indices[j][0].StartIdx
|
|
||||||
})
|
|
||||||
toRet := make([]Match, 0, len(indices))
|
|
||||||
current := indices[0]
|
|
||||||
for _, idx := range indices[1:] {
|
|
||||||
// idx doesn't overlap with current (starts after current ends), so add current to result
|
|
||||||
// and update the current.
|
|
||||||
if idx[0].StartIdx >= current[0].EndIdx {
|
|
||||||
toRet = append(toRet, current)
|
|
||||||
current = idx
|
|
||||||
} else if idx[0].EndIdx > current[0].EndIdx {
|
|
||||||
// idx overlaps, but it is longer, so update current
|
|
||||||
current = idx
|
|
||||||
}
|
|
||||||
}
|
|
||||||
// Add last state
|
|
||||||
toRet = append(toRet, current)
|
|
||||||
return toRet
|
|
||||||
}
|
|
||||||
|
|
||||||
func copyThread(to *nfaState, from nfaState) {
|
func copyThread(to *nfaState, from nfaState) {
|
||||||
to.threadGroups = append([]Group{}, from.threadGroups...)
|
to.threadGroups = append([]Group{}, from.threadGroups...)
|
||||||
}
|
}
|
||||||
@@ -223,9 +198,6 @@ func (regex Reg) FindAllSubmatch(str string) []Match {
|
|||||||
indices = append(indices, matchIdx)
|
indices = append(indices, matchIdx)
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
if len(indices) > 0 {
|
|
||||||
return pruneIndices(indices)
|
|
||||||
}
|
|
||||||
|
|
||||||
return indices
|
return indices
|
||||||
}
|
}
|
||||||
@@ -272,8 +244,6 @@ func addStateToList(str []rune, idx int, list []nfaState, state nfaState, thread
|
|||||||
// Helper for FindAllMatches. Returns whether it found a match, the
|
// Helper for FindAllMatches. Returns whether it found a match, the
|
||||||
// first Match it finds, and how far it got into the string ie. where
|
// first Match it finds, and how far it got into the string ie. where
|
||||||
// the next search should start from.
|
// the next search should start from.
|
||||||
//
|
|
||||||
// Might return duplicates or overlapping indices, so care must be taken to prune the resulting array.
|
|
||||||
func findAllSubmatchHelper(start *nfaState, str []rune, offset int, numGroups int) (bool, Match, int) {
|
func findAllSubmatchHelper(start *nfaState, str []rune, offset int, numGroups int) (bool, Match, int) {
|
||||||
// Base case - exit if offset exceeds string's length
|
// Base case - exit if offset exceeds string's length
|
||||||
if offset > len(str) {
|
if offset > len(str) {
|
||||||
@@ -282,21 +252,9 @@ func findAllSubmatchHelper(start *nfaState, str []rune, offset int, numGroups in
|
|||||||
}
|
}
|
||||||
resetThreads(start)
|
resetThreads(start)
|
||||||
|
|
||||||
// Hold a list of match indices for the current run. When we
|
|
||||||
// can no longer find a match, the match with the largest range is
|
|
||||||
// chosen as the match for the entire string.
|
|
||||||
// This allows us to pick the longest possible match (which is how greedy matching works).
|
|
||||||
// COMMENT ABOVE IS CURRENTLY NOT UP-TO-DATE
|
|
||||||
// tempIndices := newMatch(numGroups + 1)
|
|
||||||
|
|
||||||
// foundPath := false
|
|
||||||
//startIdx := offset
|
|
||||||
//endIdx := offset
|
|
||||||
currentStates := make([]nfaState, 0)
|
currentStates := make([]nfaState, 0)
|
||||||
nextStates := make([]nfaState, 0)
|
nextStates := make([]nfaState, 0)
|
||||||
// tempStates := make([]*nfaState, 0) // Used to store states that should be used in next loop iteration
|
|
||||||
i := offset // Index in string
|
i := offset // Index in string
|
||||||
//startingFrom := i // Store starting index
|
|
||||||
|
|
||||||
// If the first state is an assertion, makes sure the assertion
|
// If the first state is an assertion, makes sure the assertion
|
||||||
// is true before we do _anything_ else.
|
// is true before we do _anything_ else.
|
||||||
@@ -306,29 +264,11 @@ func findAllSubmatchHelper(start *nfaState, str []rune, offset int, numGroups in
|
|||||||
return false, []Group{}, i
|
return false, []Group{}, i
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
// Increment until we hit a character matching the start state (assuming not 0-state)
|
|
||||||
// if start.isEmpty == false {
|
|
||||||
// for i < len(str) && !start.contentContains(str, i) {
|
|
||||||
// i++
|
|
||||||
// }
|
|
||||||
// startIdx = i
|
|
||||||
// startingFrom = i
|
|
||||||
// i++ // Advance to next character (if we aren't at a 0-state, which doesn't match anything), so that we can check for transitions. If we advance at a 0-state, we will never get a chance to match the first character
|
|
||||||
// }
|
|
||||||
|
|
||||||
// start.threadGroups = newMatch(numGroups + 1)
|
|
||||||
// Check if the start state begins a group - if so, add the start index to our list
|
|
||||||
//if start.groupBegin {
|
|
||||||
// start.threadGroups[start.groupNum].StartIdx = i
|
|
||||||
// tempIndices[start.groupNum].startIdx = i
|
|
||||||
//}
|
|
||||||
|
|
||||||
start.threadGroups = newMatch(numGroups + 1)
|
start.threadGroups = newMatch(numGroups + 1)
|
||||||
start.threadGroups[0].StartIdx = i
|
start.threadGroups[0].StartIdx = i
|
||||||
currentStates = addStateToList(str, i, currentStates, *start, start.threadGroups, nil)
|
currentStates = addStateToList(str, i, currentStates, *start, start.threadGroups, nil)
|
||||||
var match Match = nil
|
var match Match = nil
|
||||||
// var isEmptyAndNoAssertion bool
|
|
||||||
// Main loop
|
|
||||||
for idx := i; idx <= len(str); idx++ {
|
for idx := i; idx <= len(str); idx++ {
|
||||||
if len(currentStates) == 0 {
|
if len(currentStates) == 0 {
|
||||||
break
|
break
|
||||||
@@ -350,76 +290,6 @@ func findAllSubmatchHelper(start *nfaState, str []rune, offset int, numGroups in
|
|||||||
nextStates = addStateToList(str, idx+1, nextStates, *currentState.next, currentState.threadGroups, nil)
|
nextStates = addStateToList(str, idx+1, nextStates, *currentState.next, currentState.threadGroups, nil)
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
// if currentState.groupBegin {
|
|
||||||
// currentState.threadGroups[currentState.groupNum].StartIdx = idx
|
|
||||||
// }
|
|
||||||
// if currentState.groupEnd {
|
|
||||||
// currentState.threadGroups[currentState.groupNum].EndIdx = idx
|
|
||||||
// }
|
|
||||||
|
|
||||||
// Alternation - enqueue left then right state, and continue
|
|
||||||
// if currentState.isAlternation {
|
|
||||||
// if currentState.isKleene { // Reverse order of adding things
|
|
||||||
// rightState := currentState.splitState
|
|
||||||
// copyThread(rightState, currentState)
|
|
||||||
// currentStates = slices.Insert(currentStates, currentStateIdx+1, *rightState)
|
|
||||||
// leftState := currentState.next
|
|
||||||
// copyThread(leftState, currentState)
|
|
||||||
// currentStates = slices.Insert(currentStates, currentStateIdx+2, *leftState)
|
|
||||||
// } else {
|
|
||||||
// leftState := currentState.next
|
|
||||||
// copyThread(leftState, currentState)
|
|
||||||
// currentStates = slices.Insert(currentStates, currentStateIdx+1, *leftState)
|
|
||||||
// rightState := currentState.splitState
|
|
||||||
// copyThread(rightState, currentState)
|
|
||||||
// currentStates = slices.Insert(currentStates, currentStateIdx+2, *rightState)
|
|
||||||
// }
|
|
||||||
// continue
|
|
||||||
// }
|
|
||||||
|
|
||||||
// Empty state - enqueue next state, do _not_ increment the SP
|
|
||||||
// if !currentState.isAlternation && currentState.isEmpty && currentState.assert == noneAssert { //&& currentState.groupBegin == false && currentState.groupEnd == false {
|
|
||||||
// isEmptyAndNoAssertion = true
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// if currentState.contentContains(str, idx) {
|
|
||||||
// foundMatch = true
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// if isEmptyAndNoAssertion || foundMatch {
|
|
||||||
// nextMatch := *(currentState.next)
|
|
||||||
// copyThread(&nextMatch, currentState)
|
|
||||||
// if currentState.groupBegin {
|
|
||||||
// // if !stateExists(currentStates, nextMatch) {
|
|
||||||
// currentStates = slices.Insert(currentStates, currentStateIdx+1, nextMatch)
|
|
||||||
// //}
|
|
||||||
// } else if currentState.groupEnd {
|
|
||||||
// if !stateExists(currentStates, nextMatch) {
|
|
||||||
// currentStates = slices.Insert(currentStates, currentStateIdx+1, nextMatch) // append(currentStates, nextMatch)
|
|
||||||
// }
|
|
||||||
// } else if currentState.assert != noneAssert {
|
|
||||||
// if !stateExists(currentStates, nextMatch) {
|
|
||||||
// currentStates = append(currentStates, nextMatch)
|
|
||||||
// }
|
|
||||||
// } else if currentState.isEmpty && !currentState.groupBegin && !currentState.groupEnd {
|
|
||||||
// if !stateExists(currentStates, nextMatch) {
|
|
||||||
// currentStates = append(currentStates, nextMatch)
|
|
||||||
// }
|
|
||||||
// } else {
|
|
||||||
// if !stateExists(nextStates, nextMatch) {
|
|
||||||
// nextStates = append(nextStates, nextMatch)
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// if currentState.isLast && len(nextStates) == 0 { // Last state reached
|
|
||||||
// currentState.threadGroups[0].EndIdx = idx
|
|
||||||
// if idx == currentState.threadGroups[0].StartIdx {
|
|
||||||
// idx += 1
|
|
||||||
// }
|
|
||||||
// return true, currentState.threadGroups, idx
|
|
||||||
// }
|
|
||||||
}
|
}
|
||||||
currentStates = append([]nfaState{}, nextStates...)
|
currentStates = append([]nfaState{}, nextStates...)
|
||||||
nextStates = nil
|
nextStates = nil
|
||||||
@@ -431,196 +301,4 @@ func findAllSubmatchHelper(start *nfaState, str []rune, offset int, numGroups in
|
|||||||
return true, match, match[0].EndIdx
|
return true, match, match[0].EndIdx
|
||||||
}
|
}
|
||||||
return false, []Group{}, i + 1
|
return false, []Group{}, i + 1
|
||||||
// zeroStates := make([]*nfaState, 0)
|
|
||||||
// // Keep taking zero-states, until there are no more left to take
|
|
||||||
// // Objective: If any of our current states have transitions to 0-states, replace them with the 0-state. Do this until there are no more transitions to 0-states, or there are no more unique 0-states to take.
|
|
||||||
// topStateItem := currentStates.peek()
|
|
||||||
// topState := topStateItem.(*priorQueueItem).state
|
|
||||||
// zeroStates, isZero := takeZeroState([]*nfaState{topState}, numGroups, i)
|
|
||||||
// tempStates = append(tempStates, zeroStates...)
|
|
||||||
// num_appended := 0
|
|
||||||
// for isZero == true {
|
|
||||||
// zeroStates, isZero = takeZeroState(tempStates, numGroups, i)
|
|
||||||
// tempStates, num_appended = uniqueAppend(tempStates, zeroStates...)
|
|
||||||
// if num_appended == 0 { // Break if we haven't appended any more unique values
|
|
||||||
// break
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
// if isZero == true {
|
|
||||||
// currentStates.Pop()
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// for _, state := range tempStates {
|
|
||||||
// heap.Push(currentStates, newPriorQueueItem(state))
|
|
||||||
// }
|
|
||||||
// tempStates = nil
|
|
||||||
//
|
|
||||||
// // Take any transitions corresponding to current character
|
|
||||||
// numStatesMatched := 0 // The number of states which had at least 1 match for this round
|
|
||||||
// assertionFailed := false // Whether or not an assertion failed for this round
|
|
||||||
// lastStateInList := false // Whether or not a last state was in our list of states
|
|
||||||
// var lastStatePtr *nfaState = nil // Pointer to the last-state, if it was found
|
|
||||||
// lastLookaroundInList := false // Whether or not a last state (that is a lookaround) was in our list of states
|
|
||||||
// for numStatesMatched == 0 && lastStateInList == false {
|
|
||||||
// if currentStates.Len() == 0 {
|
|
||||||
// break
|
|
||||||
// }
|
|
||||||
// stateItem := heap.Pop(currentStates)
|
|
||||||
// state := stateItem.(*priorQueueItem).state
|
|
||||||
// matches, numMatches := state.matchesFor(str, i)
|
|
||||||
// if numMatches > 0 {
|
|
||||||
// numStatesMatched++
|
|
||||||
// tempStates = append([]*nfaState(nil), matches...)
|
|
||||||
// foundPath = true
|
|
||||||
// for _, m := range matches {
|
|
||||||
// if m.threadGroups == nil {
|
|
||||||
// m.threadGroups = newMatch(numGroups + 1)
|
|
||||||
// }
|
|
||||||
// m.threadSP = state.threadSP + 1
|
|
||||||
// copy(m.threadGroups, state.threadGroups)
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
// if numMatches < 0 {
|
|
||||||
// assertionFailed = true
|
|
||||||
// }
|
|
||||||
// if state.isLast {
|
|
||||||
// if state.isLookaround() {
|
|
||||||
// lastLookaroundInList = true
|
|
||||||
// }
|
|
||||||
// lastStateInList = true
|
|
||||||
// lastStatePtr = state
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// if assertionFailed && numStatesMatched == 0 { // Nothing has matched and an assertion has failed
|
|
||||||
// // If I'm being completely honest, I'm not sure why I have to check specifically for a _lookaround_
|
|
||||||
// // state. The explanation below is my attempt to explain this behavior.
|
|
||||||
// // If you replace 'lastLookaroundInList' with 'lastStateInList', one of the test cases fails.
|
|
||||||
// //
|
|
||||||
// // One of the states in our list was a last state and a lookaround. In this case, we
|
|
||||||
// // don't abort upon failure of the assertion, because we have found
|
|
||||||
// // another path to a final state.
|
|
||||||
// // Even if the last state _was_ an assertion, we can use the previously
|
|
||||||
// // saved indices to find a match.
|
|
||||||
// if lastLookaroundInList {
|
|
||||||
// break
|
|
||||||
// } else {
|
|
||||||
// if i == startingFrom {
|
|
||||||
// i++
|
|
||||||
// }
|
|
||||||
// return false, []Group{}, i
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
// // Check if we can find a state in our list that is:
|
|
||||||
// // a. A last-state
|
|
||||||
// // b. Empty
|
|
||||||
// // c. Doesn't assert anything
|
|
||||||
// for _, stateItem := range *currentStates {
|
|
||||||
// s := stateItem.state
|
|
||||||
// if s.isLast && s.isEmpty && s.assert == noneAssert {
|
|
||||||
// lastStatePtr = s
|
|
||||||
// lastStateInList = true
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
// if lastStateInList && numStatesMatched == 0 { // A last-state was in the list of states. add the matchIndex to our MatchIndex list
|
|
||||||
// for j := 1; j < numGroups+1; j++ {
|
|
||||||
// tempIndices[j] = lastStatePtr.threadGroups[j]
|
|
||||||
// }
|
|
||||||
// endIdx = i
|
|
||||||
// tempIndices[0] = Group{startIdx, endIdx}
|
|
||||||
// if tempIndices[0].StartIdx == tempIndices[0].EndIdx {
|
|
||||||
// return true, tempIndices, tempIndices[0].EndIdx + 1
|
|
||||||
// } else {
|
|
||||||
// return true, tempIndices, tempIndices[0].EndIdx
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// // Check if we can find a zero-length match
|
|
||||||
// if foundPath == false {
|
|
||||||
// currentStatesList := funcMap(*currentStates, func(item *priorQueueItem) *nfaState {
|
|
||||||
// return item.state
|
|
||||||
// })
|
|
||||||
// if ok := zeroMatchPossible(str, i, numGroups, currentStatesList...); ok {
|
|
||||||
// if tempIndices[0].IsValid() == false {
|
|
||||||
// tempIndices[0] = Group{startIdx, startIdx}
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
// // If we haven't moved in the string, increment the counter by 1
|
|
||||||
// // to ensure we don't keep trying the same string over and over.
|
|
||||||
// // if i == startingFrom {
|
|
||||||
// startIdx++
|
|
||||||
// // i++
|
|
||||||
// // }
|
|
||||||
// if tempIndices.numValidGroups() > 0 && tempIndices[0].IsValid() {
|
|
||||||
// if tempIndices[0].StartIdx == tempIndices[0].EndIdx { // If we have a zero-length match, we have to shift the index at which we start. Otherwise we keep looking at the same paert of the string over and over.
|
|
||||||
// return true, tempIndices, tempIndices[0].EndIdx + 1
|
|
||||||
// } else {
|
|
||||||
// return true, tempIndices, tempIndices[0].EndIdx
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
// return false, []Group{}, startIdx
|
|
||||||
// }
|
|
||||||
// currentStates = &priorityQueue{}
|
|
||||||
// slices.Reverse(tempStates)
|
|
||||||
// for _, state := range tempStates {
|
|
||||||
// heap.Push(currentStates, newPriorQueueItem(state))
|
|
||||||
// }
|
|
||||||
// tempStates = nil
|
|
||||||
//
|
|
||||||
// i++
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// // End-of-string reached. Go to any 0-states, until there are no more 0-states to go to. Then check if any of our states are in the end position.
|
|
||||||
// // This is the exact same algorithm used inside the loop, so I should probably put it in a function.
|
|
||||||
//
|
|
||||||
// if currentStates.Len() > 0 {
|
|
||||||
// topStateItem := currentStates.peek()
|
|
||||||
// topState := topStateItem.(*priorQueueItem).state
|
|
||||||
// zeroStates, isZero := takeZeroState([]*nfaState{topState}, numGroups, i)
|
|
||||||
// tempStates = append(tempStates, zeroStates...)
|
|
||||||
// num_appended := 0 // Number of unique states addded to tempStates
|
|
||||||
// for isZero == true {
|
|
||||||
// zeroStates, isZero = takeZeroState(tempStates, numGroups, i)
|
|
||||||
// tempStates, num_appended = uniqueAppend(tempStates, zeroStates...)
|
|
||||||
// if num_appended == 0 { // Break if we haven't appended any more unique values
|
|
||||||
// break
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// for _, state := range tempStates {
|
|
||||||
// heap.Push(currentStates, newPriorQueueItem(state))
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// tempStates = nil
|
|
||||||
//
|
|
||||||
// for _, stateItem := range *currentStates {
|
|
||||||
// state := stateItem.state
|
|
||||||
// // Only add the match if the start index is in bounds. If the state has an assertion,
|
|
||||||
// // make sure the assertion checks out.
|
|
||||||
// if state.isLast && i <= len(str) {
|
|
||||||
// if state.assert == noneAssert || state.checkAssertion(str, i) {
|
|
||||||
// for j := 1; j < numGroups+1; j++ {
|
|
||||||
// tempIndices[j] = state.threadGroups[j]
|
|
||||||
// }
|
|
||||||
// endIdx = i
|
|
||||||
// tempIndices[0] = Group{startIdx, endIdx}
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// if tempIndices.numValidGroups() > 0 {
|
|
||||||
// if tempIndices[0].StartIdx == tempIndices[0].EndIdx { // If we have a zero-length match, we have to shift the index at which we start. Otherwise we keep looking at the same paert of the string over and over.
|
|
||||||
// return true, tempIndices, tempIndices[0].EndIdx + 1
|
|
||||||
// } else {
|
|
||||||
// return true, tempIndices, tempIndices[0].EndIdx
|
|
||||||
// }
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// if startIdx == startingFrom { // Increment starting index if we haven't moved in the string. Prevents us from matching the same part of the string over and over.
|
|
||||||
//
|
|
||||||
// startIdx++
|
|
||||||
// }
|
|
||||||
//
|
|
||||||
// return false, []Group{}, startIdx
|
|
||||||
}
|
}
|
||||||
|
|||||||
Reference in New Issue
Block a user