From b3dbae883ed05d3f8a9aeec01e993abd6892ea78 Mon Sep 17 00:00:00 2001
From: mpadge <mark.padgham@email.com>
Date: Tue, 22 Nov 2022 12:13:47 +0100
Subject: [PATCH] add convert_trawl_cols fn to post-process archive trawl data

---
 .pre-commit-config.yaml |  1 +
 DESCRIPTION             |  2 +-
 R/archive-trawl.R       | 49 +++++++++++++++++++++++++++++++++++++++++
 codemeta.json           |  2 +-
 4 files changed, 52 insertions(+), 2 deletions(-)

diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index b30a915..fb6229d 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -8,6 +8,7 @@ repos:
         args: [--style_pkg=spaceout, --style_fun=spaceout_style]    
         additional_dependencies:
         - ropensci-review-tools/spaceout
+        - styler
           #-   id: roxygenize
           #additional_dependencies:
           #- pkgbuild
diff --git a/DESCRIPTION b/DESCRIPTION
index 75aa58e..2103a3c 100755
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: pkgstats
 Title: Metrics of R Packages
-Version: 0.1.1.067
+Version: 0.1.1.068
 Authors@R: 
     person("Mark", "Padgham", , "mark.padgham@email.com", role = c("aut", "cre"),
            comment = c(ORCID = "0000-0003-2172-5265"))
diff --git a/R/archive-trawl.R b/R/archive-trawl.R
index 6e50947..2ee9959 100644
--- a/R/archive-trawl.R
+++ b/R/archive-trawl.R
@@ -460,3 +460,52 @@ get_pkg_version <- function (path) {
 
     c (pkg, v)
 }
+
+#' Internal function to convert columns in result of archive trawl to expected,
+#' generally numeric, types.
+#' @noRd
+convert_trawl_columns <- function (x) {
+
+    int_cols <- c (
+        "^files\\_",
+        "^loc\\_",
+        "^blank\\_",
+        "^comment\\_",
+        "^indentation$",
+        "^rel\\_space$",
+        "^num\\_",
+        "^desc\\_n\\_",
+        "^n\\_fns\\_(r|s)",
+        "^n\\_edges",
+        "^n\\_clusters"
+    )
+    num_cols <- c (
+        "^rel\\_space\\_",
+        "^nexpr$",
+        "^rel\\_space\\_",
+        "^data\\_size\\_",
+        "^n\\_fns\\_per",
+        "^npars\\_",
+        "^doclines\\_per\\_",
+        "^docchars\\_per\\_",
+        "^centrality\\_",
+        "^node\\_degree\\_",
+        "^cpl\\_instability"
+    )
+
+    cols <- names (x)
+
+    int_cols <- lapply (int_cols, function (i) grep (i, cols))
+    int_cols <- sort (unlist (int_cols))
+    for (i in int_cols) {
+        storage.mode (x [[i]]) <- "integer"
+    }
+
+    num_cols <- lapply (num_cols, function (i) grep (i, cols))
+    num_cols <- sort (unlist (num_cols))
+    for (i in num_cols) {
+        storage.mode (x [[i]]) <- "numeric"
+    }
+
+    return (x)
+}
diff --git a/codemeta.json b/codemeta.json
index 6fb0ac1..67e7b76 100644
--- a/codemeta.json
+++ b/codemeta.json
@@ -11,7 +11,7 @@
   "codeRepository": "https://github.com/ropensci-review-tools/pkgstats",
   "issueTracker": "https://github.com/ropensci-review-tools/pkgstats/issues",
   "license": "https://spdx.org/licenses/GPL-3.0",
-  "version": "0.1.1.067",
+  "version": "0.1.1.068",
   "programmingLanguage": {
     "@type": "ComputerLanguage",
     "name": "R",