From 34149980a4806f6cdef7d4cd9e8eeb5db607c448 Mon Sep 17 00:00:00 2001
From: Aadhavan Srinivasan <aadhavan@twomorecents.org>
Date: Thu, 13 Mar 2025 12:11:48 -0400
Subject: [PATCH] Started working on multiple filename arguments; prefix each
line with filename containing the line; mostly indentation changes
---
cmd/main.go | 284 ++++++++++++++++++++++++++++------------------------
1 file changed, 152 insertions(+), 132 deletions(-)
diff --git a/cmd/main.go b/cmd/main.go
index d3d7be5..2b1f7cb 100644
--- a/cmd/main.go
+++ b/cmd/main.go
@@ -6,6 +6,7 @@ import (
"fmt"
"io"
"os"
+ "slices"
"github.com/fatih/color"
@@ -25,6 +26,7 @@ func main() {
multiLineFlag := flag.Bool("t", false, "Multi-line mode. Treats newline just like any character.")
printMatchesFlag := flag.Bool("p", false, "Prints start and end index of each match. Can only be used with '-t' for multi-line mode.")
caseInsensitiveFlag := flag.Bool("i", false, "Case-insensitive. Disregard the case of all characters.")
+ recursiveFlag := flag.Bool("r", false, "Recursively search all files in the given directory.")
matchNum := flag.Int("m", 0, "Print the match with the given index. Eg. -m 3 prints the third match.")
substituteText := flag.String("s", "", "Substitute the contents of each match with the given string. Overrides -o and -v")
flag.Parse()
@@ -64,21 +66,30 @@ func main() {
// 2. Build NFA from postfix representation (Thompson's algorithm)
// 3. Run the string against the NFA
- if len(flag.Args()) < 1 || len(flag.Args()) > 2 { // flag.Args() also strips out program name
+ if len(flag.Args()) < 1 { // flag.Args() also strips out program name
+ fmt.Println("ERROR: Missing cmdline args")
+ os.Exit(22)
+ }
+ if *recursiveFlag && len(flag.Args()) < 2 { // File/Directory must be provided with '-r'
fmt.Println("ERROR: Missing cmdline args")
os.Exit(22)
}
var re string
re = flag.Args()[0]
- var inputFile *os.File
+ var inputFiles []*os.File
if len(flag.Args()) == 1 || flag.Args()[1] == "-" { // Either no file argument, or file argument is "-"
- inputFile = os.Stdin
+ if !slices.Contains(inputFiles, os.Stdin) {
+ inputFiles = append(inputFiles, os.Stdin) // os.Stdin cannot be entered more than once into the file list
+ }
} else {
- var err error
- inputFile, err = os.Open(flag.Args()[1])
- if err != nil {
- fmt.Printf("%s: No such file or directory\n", flag.Args()[1])
- os.Exit(2)
+ inputFilenames := flag.Args()[1:]
+ for _, inputFilename := range inputFilenames {
+ inputFile, err := os.Open(inputFilename)
+ if err != nil {
+ fmt.Printf("%s: No such file or directory\n", flag.Args()[1])
+ os.Exit(2)
+ }
+ inputFiles = append(inputFiles, inputFile)
}
}
@@ -86,160 +97,169 @@ func main() {
var err error
var linesRead bool // Whether or not we have read the lines in the file
lineNum := 0 // Current line number
- // Create reader for stdin and writer for stdout
- reader := bufio.NewReader(inputFile)
+ // Create writer for stdout
out := bufio.NewWriter(os.Stdout)
-
+ // Compile regex
regComp, err := reg.Compile(re, flagsToCompile...)
if err != nil {
fmt.Println(err)
return
}
- for true {
- if linesRead {
- break
- }
- if !(*multiLineFlag) {
- // Read every string from stdin until we encounter an error. If the error isn't EOF, panic.
- test_str, err = reader.ReadString('\n')
- lineNum++
- if err != nil {
+
+ for _, inputFile := range inputFiles {
+ reader := bufio.NewReader(inputFile)
+ linesRead = false
+ for true {
+ if !(*lineFlag) && (!*onlyFlag) {
+ color.New(color.FgMagenta).Fprintf(out, "%s: ", inputFile.Name()) // The filename should be printed for every line, _except_ if we're not printing every line. This is the case with the lineFlag and onlyFlag
+ }
+ if linesRead {
+ break
+ }
+ if !(*multiLineFlag) {
+ // Read every string from stdin until we encounter an error. If the error isn't EOF, panic.
+ test_str, err = reader.ReadString('\n')
+ lineNum++
+ if err != nil {
+ if err == io.EOF {
+ linesRead = true
+ } else {
+ panic(err)
+ }
+ }
+ if len(test_str) > 0 && test_str[len(test_str)-1] == '\n' {
+ test_str = test_str[:len(test_str)-1]
+ }
+ } else {
+ // Multi-line mode - read every line of input into a temp. string.
+ // test_str will contain all lines of input (including newline characters)
+ // as one string.
+ var temp string
+ for temp, err = reader.ReadString('\n'); err == nil; temp, err = reader.ReadString('\n') {
+ test_str += temp
+ }
+ // Assuming err != nil
if err == io.EOF {
+ if len(temp) > 0 {
+ test_str += temp // Add the last line (if it is non-empty)
+ }
linesRead = true
} else {
panic(err)
}
}
- if len(test_str) > 0 && test_str[len(test_str)-1] == '\n' {
- test_str = test_str[:len(test_str)-1]
- }
- } else {
- // Multi-line mode - read every line of input into a temp. string.
- // test_str will contain all lines of input (including newline characters)
- // as one string.
- var temp string
- for temp, err = reader.ReadString('\n'); err == nil; temp, err = reader.ReadString('\n') {
- test_str += temp
- }
- // Assuming err != nil
- if err == io.EOF {
- if len(temp) > 0 {
- test_str += temp // Add the last line (if it is non-empty)
+ matchIndices := make([]reg.Match, 0)
+ if matchNumFlagEnabled {
+ tmp, err := regComp.FindNthMatch(test_str, *matchNum)
+ if err == nil {
+ matchIndices = append(matchIndices, tmp)
}
- linesRead = true
} else {
- panic(err)
- }
- }
- matchIndices := make([]reg.Match, 0)
- if matchNumFlagEnabled {
- tmp, err := regComp.FindNthMatch(test_str, *matchNum)
- if err == nil {
- matchIndices = append(matchIndices, tmp)
+ matchIndices = regComp.FindAllSubmatch(test_str)
}
- } else {
- matchIndices = regComp.FindAllSubmatch(test_str)
- }
- test_str_runes := []rune(test_str) // Converting to runes preserves unicode characters
+ test_str_runes := []rune(test_str) // Converting to runes preserves unicode characters
- if *printMatchesFlag {
- // if we are in single line mode, print the line on which
- // the matches occur
- if len(matchIndices) > 0 {
- if !(*multiLineFlag) {
- fmt.Fprintf(out, "Line %d:\n", lineNum)
- }
- for _, m := range matchIndices {
- fmt.Fprintf(out, "%s\n", m.String())
- }
- err := out.Flush()
- if err != nil {
- panic(err)
+ if *printMatchesFlag {
+ // if we are in single line mode, print the line on which
+ // the matches occur
+ if len(matchIndices) > 0 {
+ if !(*multiLineFlag) {
+ fmt.Fprintf(out, "Line %d:\n", lineNum)
+ }
+ for _, m := range matchIndices {
+ fmt.Fprintf(out, "%s\n", m.String())
+ }
+ err := out.Flush()
+ if err != nil {
+ panic(err)
+ }
}
+ continue
}
- continue
- }
- // Decompose the array of matchIndex structs into a flat unique array of ints - if matchIndex is {4,7}, flat array will contain 4,5,6
- // This should make checking O(1) instead of O(n)
- indicesToPrint := new_uniq_arr[int]()
- for _, idx := range matchIndices {
- indicesToPrint.add(genRange(idx[0].StartIdx, idx[0].EndIdx)...)
- }
- // If we are inverting, then we should print the indices which _didn't_ match
- // in color.
- if *invertFlag {
- oldIndices := indicesToPrint.values()
- indicesToPrint = new_uniq_arr[int]()
- // Explanation:
- // Find all numbers from 0 to len(test_str_runes) that are NOT in oldIndices.
- // These are the values we want to print, now that we have inverted the match.
- // Re-initialize indicesToPrint and add all of these values to it.
- indicesToPrint.add(setDifference(genRange(0, len(test_str_runes)), oldIndices)...)
+ // Decompose the array of matchIndex structs into a flat unique array of ints - if matchIndex is {4,7}, flat array will contain 4,5,6
+ // This should make checking O(1) instead of O(n)
+ indicesToPrint := new_uniq_arr[int]()
+ for _, idx := range matchIndices {
+ indicesToPrint.add(genRange(idx[0].StartIdx, idx[0].EndIdx)...)
+ }
+ // If we are inverting, then we should print the indices which _didn't_ match
+ // in color.
+ if *invertFlag {
+ oldIndices := indicesToPrint.values()
+ indicesToPrint = new_uniq_arr[int]()
+ // Explanation:
+ // Find all numbers from 0 to len(test_str_runes) that are NOT in oldIndices.
+ // These are the values we want to print, now that we have inverted the match.
+ // Re-initialize indicesToPrint and add all of these values to it.
+ indicesToPrint.add(setDifference(genRange(0, len(test_str_runes)), oldIndices)...)
- }
- // If lineFlag is enabled, we should only print something if:
- // a. We are not inverting, and have at least one match on the current line
- // OR
- // b. We are inverting, and have no matches at all on the current line.
- // This checks for the inverse, and continues if it is true.
- if *lineFlag {
- if !(*invertFlag) && len(matchIndices) == 0 || *invertFlag && len(matchIndices) > 0 {
- continue
}
- }
+ // If lineFlag is enabled, we should only print something if:
+ // a. We are not inverting, and have at least one match on the current line
+ // OR
+ // b. We are inverting, and have no matches at all on the current line.
+ // This checks for the inverse, and continues if it is true.
+ if *lineFlag {
+ if !(*invertFlag) && len(matchIndices) == 0 || *invertFlag && len(matchIndices) > 0 {
+ continue
+ } else {
+ color.New(color.FgMagenta).Fprintf(out, "%s: ", inputFile.Name()) // Print filename
+ }
+ }
- // If we are substituting, we need a different behavior, as follows:
- // For every character in the test string:
- // 1. Check if the index is the start of any matchIndex
- // 2. If so, print the substitute text, and set our index to
- // the corresponding end index.
- // 3. If not, just print the character.
- if substituteFlagEnabled {
- for i := range test_str_runes {
- inMatchIndex := false
- for _, m := range matchIndices {
- if i == m[0].StartIdx {
- fmt.Fprintf(out, "%s", *substituteText)
- i = m[0].EndIdx
- inMatchIndex = true
- break
+ // If we are substituting, we need a different behavior, as follows:
+ // For every character in the test string:
+ // 1. Check if the index is the start of any matchIndex
+ // 2. If so, print the substitute text, and set our index to
+ // the corresponding end index.
+ // 3. If not, just print the character.
+ if substituteFlagEnabled {
+ for i := range test_str_runes {
+ inMatchIndex := false
+ for _, m := range matchIndices {
+ if i == m[0].StartIdx {
+ fmt.Fprintf(out, "%s", *substituteText)
+ i = m[0].EndIdx
+ inMatchIndex = true
+ break
+ }
+ }
+ if !inMatchIndex {
+ fmt.Fprintf(out, "%c", test_str_runes[i])
}
}
- if !inMatchIndex {
- fmt.Fprintf(out, "%c", test_str_runes[i])
- }
- }
- } else {
- for i, c := range test_str_runes {
- if indicesToPrint.contains(i) {
- color.New(color.FgRed).Fprintf(out, "%c", c)
- // Newline after every match - only if -o is enabled and -v is disabled.
- if *onlyFlag && !(*invertFlag) {
- for matchIdxNum, idx := range matchIndices {
- if matchIdxNum < len(matchIndices)-1 { // Only print a newline afte printing a match, if there are multiple matches on the line, and we aren't on the last one. This is because the newline that gets added at the end will take care of that.
- if i+1 == idx[0].EndIdx { // End index is one more than last index of match
- fmt.Fprintf(out, "\n")
- break
+ } else {
+ for i, c := range test_str_runes {
+ if indicesToPrint.contains(i) {
+ color.New(color.FgRed, color.Bold).Fprintf(out, "%c", c)
+ // Newline after every match - only if -o is enabled and -v is disabled.
+ if *onlyFlag && !(*invertFlag) {
+ for matchIdxNum, idx := range matchIndices {
+ if matchIdxNum < len(matchIndices)-1 { // Only print a newline afte printing a match, if there are multiple matches on the line, and we aren't on the last one. This is because the newline that gets added at the end will take care of that.
+ if i+1 == idx[0].EndIdx { // End index is one more than last index of match
+ fmt.Fprintf(out, "\n")
+ break
+ }
}
}
}
- }
- } else {
- if !(*onlyFlag) {
- fmt.Fprintf(out, "%c", c)
+ } else {
+ if !(*onlyFlag) {
+ fmt.Fprintf(out, "%c", c)
+ }
}
}
}
- }
- err = out.Flush()
- if err != nil {
- panic(err)
- }
- // If the last character in the string wasn't a newline, AND we either have don't -o set or we do (and we've matched something), then print a newline
- if (len(test_str_runes) > 0 && test_str_runes[len(test_str_runes)-1] != '\n') &&
- (!*onlyFlag || indicesToPrint.len() > 0) {
- fmt.Println()
+ err = out.Flush()
+ if err != nil {
+ panic(err)
+ }
+ // If the last character in the string wasn't a newline, AND we either have don't -o set or we do (and we've matched something), then print a newline
+ if (len(test_str_runes) > 0 && test_str_runes[len(test_str_runes)-1] != '\n') &&
+ (!*onlyFlag || indicesToPrint.len() > 0) {
+ fmt.Println()
+ }
}
}
}