--- /dev/null
+#include <err.h>
+#include <fcntl.h>
+#include <stdio.h>
+#include <unistd.h>
+
+#include <zck.h>
+
+struct chunk_stuff {
+ zckChunk *chunk;
+ size_t size;
+};
+
+static struct chunk_stuff
+get_second_chunk(struct zckCtx * const zck, const size_t chunk_count)
+{
+ size_t idx = 0;
+ zckChunk *chunk = zck_get_first_chunk(zck);
+ if (chunk == NULL)
+ errx(1, "zck_get_first_chunk() failed: %s", zck_get_error(zck));
+ printf("got first chunk %p\n", chunk);
+
+ size_t start = 0;
+ for (size_t idx = 0; idx < chunk_count; idx++) {
+ const ssize_t s_size = zck_get_chunk_size(chunk);
+ if (s_size < 0)
+ errx(1, "zck_get_chunk_size() returned invalid size %zd: %s", s_size, zck_get_error(zck));
+ const size_t size = (size_t)s_size;
+ printf("chunk %zu: start %zu size %zu\n", idx, start, size);
+
+ if (size > 0 && start > 0) {
+ printf("got it!\n");
+ return (struct chunk_stuff){
+ .chunk = chunk,
+ .size = size,
+ };
+ }
+ start += size;
+
+ chunk = zck_get_next_chunk(chunk);
+ if (chunk == NULL)
+ errx(1, "get_next_chunk() failed for %zu: %s", idx + 1, zck_get_error(zck));
+ }
+ errx(1, "Could not find the second chunk!");
+}
+
+int main(const int argc, char * const argv[])
+{
+ if (argc != 3)
+ errx(1, "Usage: chunk /path/to/file.zck /path/to/chunk.txt");
+
+ const char * const src_name = argv[1];
+ const int src_fd = open(src_name, O_RDONLY);
+ if (src_fd == -1)
+ err(1, "Could not open %s", src_name);
+
+ struct zckCtx *zck = zck_create();
+ if (zck == NULL)
+ err(1, "zck_create() failed");
+ printf("got zck context %p\n", zck);
+ if (!zck_init_read(zck, src_fd))
+ err(1, "zck_init_read() failed");
+
+ const ssize_t header_len = zck_get_header_length(zck);
+ if (header_len < 1)
+ errx(1, "Invalid header length %zd", header_len);
+ printf("header length %zd\n", header_len);
+ const ssize_t s_chunk_count = zck_get_chunk_count(zck);
+ if (s_chunk_count < 1)
+ errx(1, "Invalid chunk count %zd", s_chunk_count);
+ const size_t chunk_count = (size_t)s_chunk_count;
+ printf("chunk count %zu\n", chunk_count);
+
+ const struct chunk_stuff second = get_second_chunk(zck, chunk_count);
+ printf("got second chunk %p size %zu\n", second.chunk, second.size);
+ char * const data = malloc(second.size);
+ if (data == NULL)
+ err(1, "Could not allocate %zu bytes", second.size);
+ const ssize_t nread = zck_get_chunk_data(second.chunk, data, second.size);
+ if (nread != (ssize_t)second.size)
+ errx(1, "zck_get_chunk_data() returned %zd: %s", nread, zck_get_error(zck));
+ printf("got the data: %02x %02x %02x\n", data[0], data[1], data[2]);
+
+ zck_free(&zck);
+ if (zck != NULL)
+ errx(1, "zck_free() did not zero the pointer");
+ if (close(src_fd) == -1)
+ err(1, "Could not close %s after reading", src_name);
+
+ const char * const dst_name = argv[2];
+ printf("About to write %zu bytes to %s\n", second.size, dst_name);
+ const int dst_fd = open(dst_name, O_WRONLY | O_CREAT, 0644);
+ if (dst_fd == -1)
+ err(1, "Could not open %s for writing", dst_name);
+
+ size_t nwritten = 0;
+ while (nwritten < second.size) {
+ printf("- %zu bytes left to write\n", second.size - nwritten);
+ const ssize_t n = write(dst_fd, data + nwritten, second.size - nwritten);
+ if (n < 1)
+ err(1, "Could not write to %s", dst_name);
+ printf("- wrote %zd bytes\n", n);
+ nwritten += n;
+ }
+ if (close(dst_fd) == -1)
+ err(1, "Could not close %s after writing", dst_name);
+ printf("Whee!\n");
+ return 0;
+}
--- /dev/null
+#!/usr/bin/python3
+"""Compile a test program."""
+
+import argparse
+import dataclasses
+import os
+import pathlib
+import re
+import subprocess
+import sys
+import tempfile
+
+from typing import Callable, Dict, List
+
+
+MAGIC = bytes([0, ord("Z"), ord("C"), ord("K"), ord("1")])
+
+RE_DATA_SIZE = re.compile(
+ r""" ^
+ Data \s+ size \s* : \s*
+ (?P<size> 0 | [1-9][0-9]* )
+ \s*
+ $ """,
+ re.X,
+)
+
+RE_CHUNK_COUNT = re.compile(
+ r""" ^
+ Chunk \s+ count \s* : \s*
+ (?P<count> 0 | [1-9][0-9]* )
+ \s*
+ $ """,
+ re.X,
+)
+
+RE_CHUNKS = re.compile(
+ r""" ^
+ \s+
+ Chunk \s+
+ Checksum \s+
+ Start \s+
+ Comp \s size \s+
+ Size \s*
+ $ """,
+ re.X,
+)
+
+RE_CHUNK = re.compile(
+ r""" ^
+ \s+
+ (?P<idx> 0 | [1-9][0-9]* ) \s+
+ (?P<cksum> \S+ ) \s+
+ (?P<start> 0 | [1-9][0-9]* ) \s+
+ (?P<comp_size> 0 | [1-9][0-9]* ) \s+
+ (?P<size> 0 | [1-9][0-9]* ) \s*
+ $ """,
+ re.X,
+)
+
+
+@dataclasses.dataclass(frozen=True)
+class Config:
+ """Runtime configuration."""
+
+ # pylint: disable=too-many-instance-attributes
+
+ tempd: pathlib.Path
+ source: pathlib.Path
+ obj: pathlib.Path
+ program: pathlib.Path
+ env: Dict[str, str]
+
+ orig: pathlib.Path
+ compressed: pathlib.Path
+ uncompressed: pathlib.Path
+
+
+@dataclasses.dataclass(frozen=True)
+class Chunk:
+ """A single chunk descriptor."""
+
+ cstart: int
+ start: int
+ csize: int
+ size: int
+ cend: int
+ end: int
+
+
+def get_runenv() -> Dict[str, str]:
+ """Set up the environment for running the zchunk programs."""
+ env = dict(os.environ)
+ env["LC_ALL"] = "C.UTF-8"
+ env["LANGUAGE"] = ""
+ return env
+
+
+def parse_args(dirname: str) -> Config:
+ """Parse the command-line arguments, deduce some things."""
+ parser = argparse.ArgumentParser(prog="dictionary")
+ parser.add_argument(
+ "source", type=str, help="path to the test program source file",
+ )
+ parser.add_argument(
+ "filename", type=str, help="path to the filename to compress"
+ )
+
+ args = parser.parse_args()
+
+ tempd = pathlib.Path(dirname).absolute()
+ return Config(
+ tempd=tempd,
+ source=pathlib.Path(args.source),
+ obj=tempd / "chunk.o",
+ program=tempd / "chunk",
+ env=get_runenv(),
+ orig=pathlib.Path(args.filename).absolute(),
+ compressed=tempd / "words.txt.zck",
+ uncompressed=tempd / "chunk.txt",
+ )
+
+
+def do_compile(cfg: Config) -> None:
+ """Compile the test program."""
+ print("Fetching the C compiler flags for zck")
+ cflags = (
+ subprocess.check_output(
+ ["pkg-config", "--cflags", "zck"], shell=False, env=cfg.env
+ )
+ .decode("UTF-8")
+ .rstrip("\r\n")
+ )
+ if "\r" in cflags or "\n" in cflags:
+ sys.exit(f"`pkg-config --cflags zck` returned {cflags!r}")
+
+ if cfg.obj.exists():
+ sys.exit(f"Did not expect {cfg.obj} to exist")
+ cmd = f"cc -c -o '{cfg.obj}' {cflags} '{cfg.source}'"
+ print(f"Running {cmd!r}")
+ subprocess.check_call(cmd, shell=True, env=cfg.env)
+ if not cfg.obj.is_file():
+ sys.exit(f"{cmd!r} did not create the {cfg.obj} file")
+
+ print("Fetching the C linker flags and libraries for zck")
+ libs = (
+ subprocess.check_output(
+ ["pkg-config", "--libs", "zck"], shell=False, env=cfg.env
+ )
+ .decode("UTF-8")
+ .rstrip("\r\n")
+ )
+ if "\r" in libs or "\n" in libs:
+ sys.exit(f"`pkg-config --libs zck` returned {libs!r}")
+
+ if cfg.program.exists():
+ sys.exit(f"Did not expect {cfg.program} to exist")
+ cmd = f"cc -o '{cfg.program}' '{cfg.obj}' {libs}"
+ print(f"Running {cmd!r}")
+ subprocess.check_call(cmd, shell=True, env=cfg.env)
+ if not cfg.program.is_file():
+ sys.exit(f"{cmd!r} did not create the {cfg.program} file")
+ if not os.access(cfg.program, os.X_OK):
+ sys.exit(f"Not an executable file: {cfg.program}")
+ print(f"Looks like we got {cfg.program}")
+
+
+def do_compress(cfg: Config, orig_size: int) -> int:
+ """Compress the original file."""
+ print(f"About to compress {cfg.orig} to {cfg.compressed}")
+ if cfg.compressed.exists():
+ sys.exit(f"Did not expect {cfg.compressed} to exist")
+ subprocess.check_call(
+ ["zck", "-o", cfg.compressed, "--", cfg.orig],
+ shell=False,
+ env=cfg.env,
+ )
+ if not cfg.compressed.is_file():
+ sys.exit(f"zck did not create the {cfg.compressed} file")
+ comp_size = cfg.compressed.stat().st_size
+ print(f"{cfg.compressed} size is {comp_size} bytes long")
+ if comp_size >= orig_size:
+ sys.exit(
+ f"sizeof({cfg.compressed}) == {comp_size} : "
+ f"sizeof({cfg.orig}) == {orig_size}"
+ )
+ start = cfg.compressed.open(mode="rb").read(5)
+ print(f"{cfg.compressed} starts with {start!r}")
+ if start != MAGIC:
+ sys.exit(f"{cfg.compressed} does not start with {MAGIC!r}: {start!r}")
+
+ return comp_size
+
+
+def read_chunks(cfg: Config, orig_size: int, comp_size: int) -> Chunk:
+ """Parse the chunks of the compressed file."""
+ # pylint: disable=too-many-statements
+ output = subprocess.check_output(
+ ["zck_read_header", "-c", "--", cfg.compressed],
+ shell=False,
+ env=cfg.env,
+ ).decode("UTF-8")
+
+ params: Dict[str, int] = {}
+ chunks: List[Chunk] = []
+
+ def ignore_till_end(line: str) -> str:
+ """Ignore anything until EOF."""
+ raise NotImplementedError(line)
+
+ def parse_chunk(line: str) -> str:
+ """Parse a single chunk line."""
+ # pylint: disable=too-many-branches
+ data = RE_CHUNK.match(line)
+ if not data:
+ sys.exit(f"Unexpected line for chunk {len(chunks)}: {line!r}")
+ idx = int(data.group("idx"))
+ start = int(data.group("start"))
+ csize = int(data.group("comp_size"))
+ size = int(data.group("size"))
+
+ if idx != len(chunks):
+ sys.exit(f"Expected index {len(chunks)}: {line!r}")
+ if chunks:
+ last_chunk = chunks[-1]
+ if start != last_chunk.cend:
+ sys.exit(f"Expected start {last_chunk.cend}: {line!r}")
+ else:
+ if start != params["size_diff"]:
+ sys.exit(f"Expected start {params['size_diff']}: {line!r}")
+ last_chunk = Chunk(
+ cstart=0,
+ start=0,
+ csize=0,
+ size=0,
+ cend=params["size_diff"],
+ end=0,
+ )
+
+ next_chunk = Chunk(
+ cstart=start,
+ start=last_chunk.end,
+ csize=csize,
+ size=size,
+ cend=last_chunk.cend + csize,
+ end=last_chunk.end + size,
+ )
+ if next_chunk.cend > comp_size:
+ sys.exit(
+ f"Compressed size overflow: {next_chunk.cend} > {comp_size}"
+ )
+
+ more = idx + 1 != params["chunk_count"]
+ if more:
+ if next_chunk.end >= orig_size:
+ sys.exit(
+ f"Original size overflow: "
+ f"{next_chunk.end} >= {orig_size}"
+ )
+ else:
+ if next_chunk.cend != comp_size:
+ sys.exit(
+ f"Compressed size mismatch: "
+ f"{next_chunk.cend} != {comp_size}"
+ )
+ if next_chunk.end != orig_size:
+ sys.exit(
+ f"Original size mismatch: "
+ f"{next_chunk.end} != {orig_size}"
+ )
+
+ print(f"- appending {next_chunk!r}")
+ chunks.append(next_chunk)
+
+ if more:
+ return "parse_chunk"
+ return "ignore_till_end"
+
+ def wait_for_chunks(line: str) -> str:
+ """Wait for the 'Chunks:' line."""
+ if not RE_CHUNKS.match(line):
+ return "wait_for_chunks"
+
+ return "parse_chunk"
+
+ def wait_for_chunk_count(line: str) -> str:
+ """Wait for the 'chunk count' line."""
+ data = RE_CHUNK_COUNT.match(line)
+ if not data:
+ return "wait_for_chunk_count"
+ print(f"- got a chunk count: {data.groupdict()!r}")
+
+ count = int(data.group("count"))
+ if count < 1:
+ sys.exit(f"zck_read_header said chunk count {count}")
+ params["chunk_count"] = count
+
+ return "wait_for_chunks"
+
+ def wait_for_total_size(line: str) -> str:
+ """Wait for the 'data size' line."""
+ data = RE_DATA_SIZE.match(line)
+ if not data:
+ return "wait_for_total_size"
+ print(f"- got a size line: {data.groupdict()!r}")
+
+ size = int(data.group("size"))
+ if size < 1 or size > comp_size:
+ sys.exit(
+ f"zck_read_header said data size {size} (comp {comp_size})"
+ )
+ params["size_diff"] = comp_size - size
+
+ return "wait_for_chunk_count"
+
+ handlers: Dict[str, Callable[[str], str]] = {
+ func.__name__: func
+ for func in (
+ wait_for_total_size,
+ wait_for_chunk_count,
+ wait_for_chunks,
+ parse_chunk,
+ ignore_till_end,
+ )
+ }
+
+ handler: Callable[[str], str] = wait_for_total_size
+
+ for line in output.splitlines():
+ print(f"- read a line: {line}")
+ new_handler = handler(line)
+ assert new_handler in handlers, new_handler
+ handler = handlers[new_handler]
+
+ if handler != ignore_till_end: # pylint: disable=comparison-with-callable
+ sys.exit(f"handler is {handler!r} instead of {ignore_till_end!r}")
+
+ # Now let's find the second chunk
+ return next(chunk for chunk in chunks if chunk.start > 0)
+
+
+def run_program(cfg: Config) -> None:
+ """Run the test program, hopefully generate the chunk file."""
+ print(f"About to run {cfg.program}")
+ if cfg.uncompressed.exists():
+ sys.exit(f"Did not expect {cfg.uncompressed} to exist")
+ subprocess.check_call(
+ [cfg.program, cfg.compressed, cfg.uncompressed],
+ shell=False,
+ env=cfg.env,
+ )
+ if not cfg.uncompressed.is_file():
+ sys.exit(f"{cfg.program} did not create the {cfg.uncompressed} file")
+
+
+def compare_chunk(cfg: Config, second: Chunk, orig_size: int) -> None:
+ """Read data from the input file and the chunk."""
+ # OK, let's load it all into memory, mmkay?
+ contents = cfg.orig.read_bytes()
+ if len(contents) != orig_size:
+ sys.exit(
+ f"Could not read {orig_size} bytes from {cfg.orig}, "
+ f"read {len(contents)}"
+ )
+ chunk = cfg.uncompressed.read_bytes()
+ if len(chunk) != second.size:
+ sys.exit(
+ f"Could not read {second.size} bytes from {cfg.uncompressed}, "
+ f"read {len(chunk)}"
+ )
+
+ if contents[second.start : second.start + second.size] != chunk:
+ sys.exit("Mismatch!")
+
+
+def main() -> None:
+ """Parse arguments, compile a program, compress a file, test it."""
+ with tempfile.TemporaryDirectory() as dirname:
+ print(f"Using temporary directory {dirname}")
+ cfg = parse_args(dirname)
+ do_compile(cfg)
+ orig_size = cfg.orig.stat().st_size
+ print(f"Original file size: {orig_size}")
+ comp_size = do_compress(cfg, orig_size)
+ second_chunk = read_chunks(cfg, orig_size, comp_size)
+ run_program(cfg)
+ compare_chunk(cfg, second_chunk, orig_size)
+ print("Seems fine!")
+
+
+if __name__ == "__main__":
+ main()
projects / zchunk.git / commitdiff