From 0b4f0d6abc1b3b59afd781d147af2f60be556603 Mon Sep 17 00:00:00 2001
From: Jeremy Singer-Vine <jsvine@gmail.com>
Date: Wed, 2 Mar 2016 19:59:41 -0500
Subject: [PATCH] Move cluster functions to helpers.py

Not proper utils, and not pandas-dependent.
---
 pdfplumber/helpers.py | 28 ++++++++++++++++++++++++++++
 pdfplumber/utils.py   | 38 +++-----------------------------------
 2 files changed, 31 insertions(+), 35 deletions(-)
 create mode 100644 pdfplumber/helpers.py

diff --git a/pdfplumber/helpers.py b/pdfplumber/helpers.py
new file mode 100644
index 0000000..1584c06
--- /dev/null
+++ b/pdfplumber/helpers.py
@@ -0,0 +1,28 @@
+import itertools
+def cluster_list(xs, tolerance=0):
+    if tolerance == 0: return [ [x] for x in sorted(xs) ]
+    if len(xs) < 2: return [ [x] for x in sorted(xs) ]
+    groups = []
+    xs = list(sorted(xs))
+    current_group = [xs[0]]
+    last = xs[0]
+    for x in xs[1:]:
+        if x <= (last + tolerance):
+            current_group.append(x)
+        else:
+            groups.append(current_group)
+            current_group = [x]
+        last = x
+    groups.append(current_group)
+    return groups
+
+def make_cluster_dict(values, tolerance):
+    clusters = cluster_list(set(values), tolerance)
+
+    nested_tuples = [ [ (val, i) for val in value_cluster ]
+        for i, value_cluster in enumerate(clusters) ]
+
+    cluster_dict = dict(itertools.chain(*nested_tuples))
+    return cluster_dict 
+
+
diff --git a/pdfplumber/utils.py b/pdfplumber/utils.py
index df34b0d..6cd313b 100644
--- a/pdfplumber/utils.py
+++ b/pdfplumber/utils.py
@@ -1,37 +1,5 @@
 import pandas as pd
-import itertools
-
-def compatible_iter(thing):
-    if hasattr(thing, "iterrows"):
-        return thing.iterrows()
-    else:
-        return enumerate(thing)
-
-def cluster_list(xs, tolerance=0):
-    if tolerance == 0: return [ [x] for x in sorted(xs) ]
-    if len(xs) < 2: return [ [x] for x in sorted(xs) ]
-    groups = []
-    xs = list(sorted(xs))
-    current_group = [xs[0]]
-    last = xs[0]
-    for x in xs[1:]:
-        if x <= (last + tolerance):
-            current_group.append(x)
-        else:
-            groups.append(current_group)
-            current_group = [x]
-        last = x
-    groups.append(current_group)
-    return groups
-
-def make_cluster_dict(values, tolerance):
-    clusters = cluster_list(set(values), tolerance)
-
-    nested_tuples = [ [ (val, i) for val in value_cluster ]
-        for i, value_cluster in enumerate(clusters) ]
-
-    cluster_dict = dict(itertools.chain(*nested_tuples))
-    return cluster_dict 
+from pdfplumber import helpers
 
 def collate_chars(chars, x_tolerance=0, y_tolerance=0):
     using_pandas = isinstance(chars, pd.DataFrame)
@@ -50,7 +18,7 @@ def collate_chars(chars, x_tolerance=0, y_tolerance=0):
             coll += char["text"]
         return coll
 
-    doctop_clusters = make_cluster_dict(_chars["doctop"], y_tolerance)
+    doctop_clusters = helpers.make_cluster_dict(_chars["doctop"], y_tolerance)
     _chars["doctop_cluster"] = _chars["doctop"].apply(doctop_clusters.get)
     dc_grp = _chars.sort_values("doctop_cluster").groupby("doctop_cluster")
     coll = "\n".join(dc_grp.apply(collate_line))
@@ -114,7 +82,7 @@ def extract_columns(chars,
 
     collator = lambda x: collate_chars(x, x_tolerance=x_tolerance, y_tolerance=y_tolerance)
 
-    doctop_clusters = make_cluster_dict(_chars["doctop"], y_tolerance)
+    doctop_clusters = helpers.make_cluster_dict(_chars["doctop"], y_tolerance)
     _chars["doctop_cluster"] = _chars["doctop"].apply(doctop_clusters.get)
 
     collated = _chars.groupby([ "doctop_cluster", "column" ])\