| 1 | @[deprecated: 'If you use it, create a local module named `string_reader`, and copy https://github.com/vlang/v/blob/f24d49259db5d2d7211e79bcf6b6f507d22847f3/vlib/io/string_reader/string_reader.v there.'] |
| 2 | @[deprecated_after: '2026-01-06'] |
| 3 | module string_reader |
| 4 | |
| 5 | import io |
| 6 | import strings |
| 7 | |
| 8 | @[params] |
| 9 | pub struct StringReaderParams { |
| 10 | pub: |
| 11 | // the reader interface |
| 12 | reader ?io.Reader |
| 13 | // initialize the builder with this source string |
| 14 | source ?string |
| 15 | // if no source is given the string builder is initialized with this size |
| 16 | initial_size int |
| 17 | } |
| 18 | |
| 19 | // StringReader is able to read data from a Reader interface and/or source string to a dynamically |
| 20 | // growing buffer using a string builder. Unlike the BufferedReader, StringReader will |
| 21 | // keep the entire contents of the buffer in memory, allowing the incoming data to be reused |
| 22 | // and read in an efficient matter. The StringReader will not set a maximum capacity to the string |
| 23 | // builders buffer and could grow very large. |
| 24 | pub struct StringReader { |
| 25 | mut: |
| 26 | reader ?io.Reader |
| 27 | offset int // current offset in the buffer |
| 28 | pub mut: |
| 29 | end_of_stream bool // whether we reached the end of the upstream reader |
| 30 | builder strings.Builder |
| 31 | } |
| 32 | |
| 33 | // new creates a new StringReader and sets the string builder size to `initial_size`. |
| 34 | // If a source |
| 35 | pub fn StringReader.new(params StringReaderParams) StringReader { |
| 36 | mut r := StringReader{ |
| 37 | reader: params.reader |
| 38 | } |
| 39 | |
| 40 | if source := params.source { |
| 41 | r.builder = strings.new_builder(source.len) |
| 42 | r.builder.write_string(source) |
| 43 | if params.reader == none { |
| 44 | // There is no upstream reader; only the preloaded buffer can be consumed. |
| 45 | r.end_of_stream = true |
| 46 | } |
| 47 | } else { |
| 48 | r.builder = strings.new_builder(params.initial_size) |
| 49 | } |
| 50 | |
| 51 | return r |
| 52 | } |
| 53 | |
| 54 | // needs_fill returns whether the buffer needs refilling |
| 55 | pub fn (r StringReader) needs_fill() bool { |
| 56 | return r.offset >= r.builder.len |
| 57 | } |
| 58 | |
| 59 | // needs_fill_until returns whether the buffer needs refilling in order to read |
| 60 | // `n` bytes |
| 61 | pub fn (r StringReader) needs_fill_until(n int) bool { |
| 62 | return r.offset + n > r.builder.len |
| 63 | } |
| 64 | |
| 65 | // fill_bufer tries to read data into the buffer until either a 0 length read or if read_to_end_of_stream |
| 66 | // is true then the end of the stream. It returns the number of bytes read |
| 67 | pub fn (mut r StringReader) fill_buffer(read_till_end_of_stream bool) !int { |
| 68 | if r.end_of_stream { |
| 69 | return io.Eof{} |
| 70 | } |
| 71 | |
| 72 | mut reader := r.reader or { return error('reader is not set') } |
| 73 | |
| 74 | start := r.builder.len |
| 75 | mut end := start |
| 76 | |
| 77 | // make sure there is enough room in the string builder |
| 78 | unsafe { r.builder.grow_len(io.read_all_len) } |
| 79 | defer { |
| 80 | // shrink the length of the buffer to the total of bytes read |
| 81 | r.builder.go_back(r.builder.len - end) |
| 82 | } |
| 83 | |
| 84 | for { |
| 85 | read := reader.read(mut r.builder[end..]) or { |
| 86 | r.end_of_stream = true |
| 87 | break |
| 88 | } |
| 89 | end += read |
| 90 | |
| 91 | if !read_till_end_of_stream || read == 0 { |
| 92 | break |
| 93 | } else if r.builder.len == end { |
| 94 | unsafe { r.builder.grow_len(io.read_all_grow_len) } |
| 95 | } |
| 96 | } |
| 97 | |
| 98 | if end == start { |
| 99 | return io.Eof{} |
| 100 | } |
| 101 | |
| 102 | return end - start |
| 103 | } |
| 104 | |
| 105 | // fill_buffer_until tries read `n` amount of bytes from the reader into the buffer |
| 106 | // and returns the actual number of bytes read |
| 107 | pub fn (mut r StringReader) fill_buffer_until(n int) !int { |
| 108 | if r.end_of_stream { |
| 109 | return io.Eof{} |
| 110 | } |
| 111 | |
| 112 | mut reader := r.reader or { return error('reader is not set') } |
| 113 | |
| 114 | start := r.builder.len |
| 115 | // make sure there is enough room in the string builder |
| 116 | if n > io.read_all_len { |
| 117 | unsafe { r.builder.grow_len(io.read_all_len) } |
| 118 | } else { |
| 119 | unsafe { r.builder.grow_len(n) } |
| 120 | } |
| 121 | |
| 122 | mut end := start |
| 123 | defer { |
| 124 | // shrink the length of the buffer to the total of bytes read |
| 125 | r.builder.go_back(r.builder.len - end) |
| 126 | } |
| 127 | for { |
| 128 | read := reader.read(mut r.builder[end..]) or { |
| 129 | r.end_of_stream = true |
| 130 | break |
| 131 | } |
| 132 | end += read |
| 133 | |
| 134 | if read == 0 || end - start == n { |
| 135 | break |
| 136 | } else if r.builder.len == end { |
| 137 | remaining := n - (end - start) |
| 138 | if remaining > io.read_all_grow_len { |
| 139 | unsafe { r.builder.grow_len(io.read_all_grow_len) } |
| 140 | } else { |
| 141 | unsafe { r.builder.grow_len(remaining) } |
| 142 | } |
| 143 | } |
| 144 | } |
| 145 | |
| 146 | if end == start { |
| 147 | return io.Eof{} |
| 148 | } |
| 149 | return end - start |
| 150 | } |
| 151 | |
| 152 | // read_all_bytes reads all bytes from a reader until either a 0 length read or if read_to_end_of_stream |
| 153 | // is true then the end of the stream. It returns a copy of the read data |
| 154 | pub fn (mut r StringReader) read_all_bytes(read_till_end_of_stream bool) ![]u8 { |
| 155 | start := r.offset |
| 156 | // ignore Eof error from fill buffer |
| 157 | r.fill_buffer(read_till_end_of_stream) or {} |
| 158 | r.offset = r.builder.len |
| 159 | // check if there was still data in the buffer, but the reader has reached its end of stream |
| 160 | if start == r.offset { |
| 161 | return io.Eof{} |
| 162 | } |
| 163 | |
| 164 | return r.get_part(start, r.offset - start)! |
| 165 | } |
| 166 | |
| 167 | // read_all reads all bytes from a reader until either a 0 length read or if read_to_end_of_stream |
| 168 | // is true then the end of the stream. It produces a string from the read data |
| 169 | pub fn (mut r StringReader) read_all(read_till_end_of_stream bool) !string { |
| 170 | buf := r.read_all_bytes(read_till_end_of_stream)! |
| 171 | return unsafe { tos(buf.data, buf.len) } |
| 172 | } |
| 173 | |
| 174 | // read_bytes tries to read n amount of bytes from the reader |
| 175 | pub fn (mut r StringReader) read_bytes(n int) ![]u8 { |
| 176 | start := r.offset |
| 177 | |
| 178 | if r.needs_fill_until(n) { |
| 179 | actual_read := r.fill_buffer_until(n - (r.builder.len - r.offset))! |
| 180 | r.offset += actual_read |
| 181 | } else { |
| 182 | r.offset += n |
| 183 | } |
| 184 | |
| 185 | return r.get_part(start, r.offset - start)! |
| 186 | } |
| 187 | |
| 188 | // read_bytes tries to read `n` amount of bytes from the reader and produces a string |
| 189 | // from the read data |
| 190 | pub fn (mut r StringReader) read_string(n int) !string { |
| 191 | buf := r.read_bytes(n)! |
| 192 | return unsafe { tos(buf.data, buf.len) } |
| 193 | } |
| 194 | |
| 195 | // read implements the Reader interface |
| 196 | pub fn (mut r StringReader) read(mut buf []u8) !int { |
| 197 | if buf.len == 0 { |
| 198 | return 0 |
| 199 | } |
| 200 | |
| 201 | available := r.builder.len - r.offset |
| 202 | if available < buf.len && !r.end_of_stream { |
| 203 | if r.reader != none { |
| 204 | missing := buf.len - available |
| 205 | r.fill_buffer_until(missing) or { |
| 206 | if err !is io.Eof { |
| 207 | return err |
| 208 | } |
| 209 | } |
| 210 | } |
| 211 | } |
| 212 | |
| 213 | available_after_fill := r.builder.len - r.offset |
| 214 | if available_after_fill == 0 { |
| 215 | if r.reader == none && !r.end_of_stream { |
| 216 | return error('reader is not set') |
| 217 | } |
| 218 | return io.Eof{} |
| 219 | } |
| 220 | |
| 221 | read := if available_after_fill < buf.len { available_after_fill } else { buf.len } |
| 222 | start := r.offset |
| 223 | r.offset += read |
| 224 | |
| 225 | copy(mut buf, r.builder[start..start + read]) |
| 226 | return read |
| 227 | } |
| 228 | |
| 229 | // read_line attempts to read a line from the reader. |
| 230 | // It reads until it finds the specified delimiter or the end of stream. |
| 231 | // The returned string does not include the delimiter. |
| 232 | // When the delimiter is `\n`, a preceding `\r` is treated as part of CRLF |
| 233 | // and is also omitted from the returned string. |
| 234 | @[direct_array_access] |
| 235 | pub fn (mut r StringReader) read_line(config io.BufferedReadLineConfig) !string { |
| 236 | if r.end_of_stream && r.needs_fill() { |
| 237 | return io.Eof{} |
| 238 | } |
| 239 | |
| 240 | start := r.offset |
| 241 | for { |
| 242 | if r.needs_fill() { |
| 243 | r.fill_buffer(false) or { |
| 244 | // we are at the end of the stream |
| 245 | if r.offset == start { |
| 246 | return io.Eof{} |
| 247 | } |
| 248 | return r.get_string_part(start, r.offset - start)! |
| 249 | } |
| 250 | } |
| 251 | // try to find a newline character |
| 252 | mut i := r.offset |
| 253 | for ; i < r.builder.len; i++ { |
| 254 | c := r.builder[i] |
| 255 | if c == config.delim { |
| 256 | // great, we hit something |
| 257 | // do some checking for whether we hit \r\n or just \n |
| 258 | mut x := i |
| 259 | if i > start && config.delim == `\n` && r.builder[i - 1] == `\r` { |
| 260 | x-- |
| 261 | } |
| 262 | r.offset = i + 1 |
| 263 | return r.get_string_part(start, x - start)! |
| 264 | } |
| 265 | } |
| 266 | r.offset = i |
| 267 | } |
| 268 | |
| 269 | return io.Eof{} |
| 270 | } |
| 271 | |
| 272 | // write implements the Writer interface |
| 273 | pub fn (mut r StringReader) write(buf []u8) !int { |
| 274 | return r.builder.write(buf)! |
| 275 | } |
| 276 | |
| 277 | // get_data returns a copy of the buffer |
| 278 | @[inline] |
| 279 | pub fn (r StringReader) get_data() []u8 { |
| 280 | unsafe { |
| 281 | mut x := malloc_noscan(r.builder.len) |
| 282 | vmemcpy(x, &u8(r.builder.data), r.builder.len) |
| 283 | return x.vbytes(r.builder.len) |
| 284 | } |
| 285 | } |
| 286 | |
| 287 | // get get_part returns a copy of a part of the buffer from `start` till `start` + `n` |
| 288 | pub fn (r StringReader) get_part(start int, n int) ![]u8 { |
| 289 | if start + n > r.builder.len { |
| 290 | return io.Eof{} |
| 291 | } |
| 292 | |
| 293 | unsafe { |
| 294 | mut x := malloc_noscan(n) |
| 295 | vmemcpy(x, &u8(r.builder.data) + start, n) |
| 296 | return x.vbytes(n) |
| 297 | } |
| 298 | } |
| 299 | |
| 300 | // get_string produces a string from all the bytes in the buffer |
| 301 | @[inline] |
| 302 | pub fn (r StringReader) get_string() string { |
| 303 | return r.builder.spart(0, r.builder.len) |
| 304 | } |
| 305 | |
| 306 | // get_string_part produces a string from `start` till `start` + `n` of the buffer |
| 307 | pub fn (r StringReader) get_string_part(start int, n int) !string { |
| 308 | if start + n > r.builder.len { |
| 309 | return io.Eof{} |
| 310 | } |
| 311 | |
| 312 | return r.builder.spart(start, n) |
| 313 | } |
| 314 | |
| 315 | // flush clears the stringbuilder and returns the resulting string and the stringreaders |
| 316 | // offset is reset to 0 |
| 317 | pub fn (mut r StringReader) flush() string { |
| 318 | r.offset = 0 |
| 319 | return r.builder.str() |
| 320 | } |
| 321 | |
| 322 | // free frees the memory block used for the string builders buffer, |
| 323 | // a new string builder with size 0 is initialized and the stringreaders offset is reset to 0 |
| 324 | @[unsafe] |
| 325 | pub fn (mut r StringReader) free() { |
| 326 | unsafe { r.builder.free() } |
| 327 | r.builder = strings.new_builder(0) |
| 328 | r.offset = 0 |
| 329 | } |
| 330 | |