Minor improvements

Author: vigna

README.md

@@ -11,10 +11,10 @@ framework](https://webgraph.di.unimi.it/) for graph compression.
 - Compressed graph representation (start from here):
   [`webgraph`](https://crates.io/crates/webgraph) ([repo](/webgraph))
 
-- Algorithms: [`webgraph-algo`](https://crates.io/crates/webgraph_algo)
+- Algorithms: [`webgraph-algo`](https://crates.io/crates/webgraph-algo)
   ([repo](/algo))
 
-- CLI commands: [`webgraph-cli`](https://crates.io/crates/webgraph_cli)
+- CLI commands: [`webgraph-cli`](https://crates.io/crates/webgraph-cli)
   ([repo](/cli))
 
 ## Acknowledgments

algo/README.md

@@ -33,6 +33,10 @@ compression.
 - **[Layered Label Propagation]** (LLP): Fast community detection algorithm for
   large graphs
 
+### Ranking
+
+- **[PageRank]**: Fast parallel computation of PageRank.
+
 ## CLI Integration
 
 Many algorithms can also be accessed through the [command-line interface].
@@ -58,3 +62,4 @@ Union nor the Italian MUR can be held responsible for them.
 [Layered Label Propagation]: https://docs.rs/webgraph-algo/latest/webgraph_algo/llp/index.html
 [command-line interface]: https://docs.rs/webgraph-cli/latest/index.html
 [WebGraph framework]: https://webgraph.di.unimi.it/
+[PageRank]: https://docs.rs/webgraph-algo/latest/webgraph_algo/rank/pagerank/index.html

algo/src/llp/preds.rs

@@ -33,6 +33,7 @@ use std::fmt::Display;
 #[doc(hidden)]
 /// This structure is passed to stopping predicates to provide the information
 /// that is needed to evaluate them.
+#[derive(Debug)]
 pub struct PredParams {
     pub num_nodes: usize,
     pub num_arcs: u64,

algo/src/rank/pagerank.rs

@@ -164,6 +164,7 @@ pub mod preds {
     #[doc(hidden)]
     /// This structure is passed to stopping predicates to provide the
     /// information that is needed to evaluate them.
+    #[derive(Debug)]
     pub struct PredParams {
         pub iteration: usize,
         pub norm_delta: f64,
@@ -372,6 +373,18 @@ pub struct PageRank<'a, G: RandomAccessGraph + Sync> {
     iteration: usize,
 }
 
+impl<G: RandomAccessGraph + Sync> std::fmt::Debug for PageRank<'_, G> {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        f.debug_struct("PageRank")
+            .field("alpha", &self.alpha)
+            .field("mode", &self.mode)
+            .field("granularity", &self.granularity)
+            .field("norm_delta", &self.norm_delta)
+            .field("iteration", &self.iteration)
+            .finish_non_exhaustive()
+    }
+}
+
 impl<'a, G: RandomAccessGraph + Sync> PageRank<'a, G> {
     /// Creates a new PageRank computation.
     ///

cli/README.md

@@ -30,7 +30,7 @@ implemented as a submodule.
 - `build`: builds accessory graph data structures;
 - `check`: checks coherence of graph files;
 - `from`: ingests data into graphs;
-- `perm`: creates and manipulate permutations;
+- `perm`: creates and manipulates permutations;
 - `run`: runs algorithms on graph;
 - `to`: converts graphs between representations;
 - `transform`: applies transformations to graphs.

cli/src/lib.rs

@@ -67,7 +67,7 @@ build info: built on {} for {} with {}",
 ///
 /// It is used to implement [`ValueEnum`] here instead of in [`dsi_bitstream`].
 ///
-/// For CLI ergonomics and compatibility, this codes must be the same as those
+/// For CLI ergonomics and compatibility, these codes must be the same as those
 /// appearing in [`CompFlags::code_from_str`].
 pub enum PrivCode {
     Unary,
@@ -523,7 +523,7 @@ impl IntVectorFormat {
     /// be computed from the data.
     ///
     /// This helper method is available only on 64-bit architectures as Java's format
-    /// uses of 64-bit integers.
+    /// uses 64-bit integers.
     pub fn store_usizes(
         &self,
         path: impl AsRef<Path>,

webgraph/README.md

@@ -72,7 +72,7 @@ bitstream containing pointers into the graph bitstream).
 
 As a first step, if you need random access to the successors of a node, you need
 to build an [Elias–Fano] representation of the offsets (this part can be skipped
-if you just need sequential access). There is a [CLI command `webgraph`] with many
+if you just need sequential access). There is a [command-line interface] with many
 subcommands, among which `build`, and `webgraph build ef BASENAME` will build
 the representation for you, serializing it with [ε-serde] in a file
 named `BASENAME.ef`.
@@ -160,7 +160,7 @@ very large graphs, the mapping might not fit in RAM. For example,
 ```bash
 echo -e "a\tb\nb\tc\nc\ta" > graph.tsv
 # convert to bvgraph
-cat graph.tsv | cargo run --release from arcs --label graph
+cat graph.tsv | cargo run --release from arcs --labels graph
 ```
 
 The graph can be converted back in the arcs format using the `to arcs` command.
@@ -206,7 +206,7 @@ Union nor the Italian MUR can be held responsible for them.
 [iteration]: https://docs.rs/webgraph/latest/webgraph/traits/labels/trait.SequentialLabeling.html#method.iter
 [retrieve the successors of a node]: https://docs.rs/webgraph/latest/webgraph/traits/graph/trait.RandomAccessGraph.html#method.successors
 [LAW web site]: http://law.di.unimi.it/
-[Elias–Fano]: sux::dict::EliasFano
+[Elias–Fano]: https://docs.rs/sux/latest/sux/dict/elias_fano/struct.EliasFano.html
 [WebGraph framework]: https://webgraph.di.unimi.it/
 [ε-serde]: https://crates.io/crates/epserde/
 [`for_`]: https://docs.rs/lender/latest/lender/macro.for_.html

webgraph/src/graphs/bvgraph/comp/bvcompz.rs

@@ -17,9 +17,9 @@ use lender::prelude::*;
 
 /// An Entry for the table used to save the intermediate computation
 /// of the dynamic algorithm to select the best references.
-/// It represents if a reference to a node, with a know amount of previous
-/// references chain length, is chosen and how much less it costs to all his
-/// referent with respect to compress the node without any selected reference.
+/// It represents if a reference to a node, with a known amount of previous
+/// references chain length, is chosen and how much less it costs to all its
+/// referents with respect to compress the node without any selected reference.
 #[derive(Default, Clone)]
 struct ReferenceTableEntry {
     saved_cost: f32,

webgraph/src/graphs/random/er.rs

@@ -87,7 +87,9 @@ impl<'succ> Lending<'succ> for NodeLabels {
     type Lend = (usize, <Self as NodeLabelsLender<'succ>>::IntoIterator);
 }
 
+#[derive(Debug)]
 pub struct IntoSucc(Vec<usize>);
+#[derive(Debug)]
 pub struct Succ(IntoIter<usize>);
 
 impl Iterator for Succ {

webgraph/src/traits/graph.rs

@@ -16,13 +16,13 @@ as successors. Analogously, a [random-access graph](RandomAccessGraph) is simply
 [`RandomAccessLabeling`] extending a [`SequentialLabeling`] whose [`Label`](SequentialLabeling::Label) is `usize`.
 To access the successors of a node, however, you must use
 [`RandomAccessGraph::successors`], which delegates to [`labels`](RandomAccessLabeling::labels):
-the latter method is overridden on purpose make its usage on graphs impossible.
+the latter method is overridden on purpose to make its usage on graphs impossible.
 
 In the same vein, a [sequential graph with labels](LabeledSequentialGraph) of type `L` is a
 [`SequentialLabeling`] whose [`Label`](SequentialLabeling::Label) is `(usize, L)`
 and a [random-access graph with labels](LabeledRandomAccessGraph) is a
 [`RandomAccessLabeling`] extending a [`SequentialLabeling`] whose [`Label`](SequentialLabeling::Label) is `(usize, L)`.
-Also in this case, access the successors of a node and their labels, you must use
+Also in this case, to access the successors of a node and their labels, you must use
 [`LabeledRandomAccessGraph::successors`].
 
 Finally, the [zipping of a graph and a labeling](Zip) implements the