osmpbf-parser
parse open street map protobuf files and scan for records with explicit file offsets
This crate differs from osmpbf mainly in that you get explicit u64 file offsets for jumping and
scanning and can bring your own concurrency strategy by creating many file handles and parser
instances for the same file to fan out operations in parallel.
This example loops over all the blobs in an osmpbf file and prints diagnostic info such as the
element type (node, way, or relation), the file offset and byte length of each blob, the number of
items, and the range of ids.
```rust,norun
use std::fs::File;
use osmpbfparser::{Parser,Element};
type Error = Box;
fn main() -> Result<(),Error> {
let args = std::env::args().collect::>();
let h = File::open(&args[1])?;
let filelen = h.metadata()?.len();
let mut opd = Parser::new(Box::new(h));
let mut offset = 0;
while offset < filelen {
let (bytelen,items) = opd.read(offset)?;
let mut etype = "";
let mut minid = i64::MAX;
let mut maxid = i64::MIN;
for item in items.iter() {
match item {
Element::Node(node) => {
etype = "node";
minid = node.id.min(minid);
maxid = node.id.max(maxid);
},
Element::Way(way) => {
etype = "way";
minid = way.id.min(minid);
maxid = way.id.max(maxid);
},
Element::Relation(relation) => {
etype = "relation";
minid = relation.id.min(minid);
maxid = relation.id.max(maxid);
},
}
}
if !items.isempty() {
println!["{}: offset={} bytelen={} items.len()={} id range {}..{}",
etype, offset, bytelen, items.len(), minid, maxid];
}
offset += byte_len;
}
Ok(())
}
```
There are more explicit routines to read the file block, blob header, and blob without decoding.