[FEA] Separate each ON join by newline (#80)

This closes #76
PabloRMira · Dec 20, 2020 · 0ad891d · 0ad891d
1 parent 9534d91
commit 0ad891d
Show file tree

Hide file tree

Showing 3 changed files with 160 additions and 27 deletions.
diff --git a/docs/core.html b/docs/core.html
@@ -71,7 +71,7 @@ <h2 id="General-formatting">General formatting<a class="anchor-link" href="#Gene
 <span class="s2">table2 as b -- and here a comment</span>
 <span class="s2">    on a.asdf = b.asdf  /* joiN this way */</span>
 <span class="s2">    inner join table3 as c</span>
-<span class="s2">on a.asdf=c.asdf</span>
+<span class="s2">on a.asdf=c.asdf and a.qwer= b.qwer</span>
 <span class="s2">whEre a.asdf= 1 -- comment this</span>
 <span class="s2">anD b.qwer =2 and a.asdf&lt;=1 --comment that</span>
 <span class="s2">or b.qwer&gt;=5</span>
@@ -116,7 +116,8 @@ <h2 id="General-formatting">General formatting<a class="anchor-link" href="#Gene
 <span class="s2">    LEFT JOIN table2 as b -- and here a comment</span>
 <span class="s2">        ON a.asdf = b.asdf /* joiN this way */</span>
 <span class="s2">    INNER JOIN table3 as c</span>
-<span class="s2">        ON a.asdf = c.asdf</span>
+<span class="s2">        ON a.asdf = c.asdf and</span>
+<span class="s2">           a.qwer = b.qwer</span>
 <span class="s2">WHERE  a.asdf = 1 -- comment this</span>
 <span class="s2">   and b.qwer = 2</span>
 <span class="s2">   and a.asdf &lt;= 1 --comment that</span>
@@ -2474,6 +2475,84 @@ <h4 id="format_on" class="doc_header"><code>format_on</code><a href="https://git
 </div>
 </div>
 
+</div>
+    {% endraw %}
+
+    {% raw %}
+
+<div class="cell border-box-sizing code_cell rendered">
+<div class="input">
+
+<div class="inner_cell">
+    <div class="input_area">
+<div class=" highlight hl-ipython3"><pre><span></span><span class="n">assert_and_print</span><span class="p">(</span>
+    <span class="n">format_on</span><span class="p">(</span><span class="s2">&quot;on a.asdf =b.asdf and a.qwer=b.qwer&quot;</span><span class="p">),</span> 
+<span class="sd">&quot;&quot;&quot;</span>
+<span class="sd">        on a.asdf = b.asdf and</span>
+<span class="sd">           a.qwer = b.qwer</span>
+<span class="sd">&quot;&quot;&quot;</span><span class="o">.</span><span class="n">strip</span><span class="p">(</span><span class="s2">&quot;</span><span class="se">\n</span><span class="s2">&quot;</span><span class="p">)</span>
+<span class="p">)</span>
+</pre></div>
+
+    </div>
+</div>
+</div>
+
+<div class="output_wrapper">
+<div class="output">
+
+<div class="output_area">
+
+<div class="output_subarea output_stream output_stdout output_text">
+<pre>        on a.asdf = b.asdf and
+           a.qwer = b.qwer
+</pre>
+</div>
+</div>
+
+</div>
+</div>
+
+</div>
+    {% endraw %}
+
+    {% raw %}
+
+<div class="cell border-box-sizing code_cell rendered">
+<div class="input">
+
+<div class="inner_cell">
+    <div class="input_area">
+<div class=" highlight hl-ipython3"><pre><span></span><span class="n">assert_and_print</span><span class="p">(</span>
+    <span class="n">format_on</span><span class="p">(</span><span class="s2">&quot;on a.asdf =b.asdf and a.qwer=b.qwer or a.qwer2= b.qwer2&quot;</span><span class="p">),</span> 
+<span class="sd">&quot;&quot;&quot;</span>
+<span class="sd">        on a.asdf = b.asdf and</span>
+<span class="sd">           a.qwer = b.qwer or</span>
+<span class="sd">           a.qwer2 = b.qwer2</span>
+<span class="sd">&quot;&quot;&quot;</span><span class="o">.</span><span class="n">strip</span><span class="p">(</span><span class="s2">&quot;</span><span class="se">\n</span><span class="s2">&quot;</span><span class="p">)</span>
+<span class="p">)</span>
+</pre></div>
+
+    </div>
+</div>
+</div>
+
+<div class="output_wrapper">
+<div class="output">
+
+<div class="output_area">
+
+<div class="output_subarea output_stream output_stdout output_text">
+<pre>        on a.asdf = b.asdf and
+           a.qwer = b.qwer or
+           a.qwer2 = b.qwer2
+</pre>
+</div>
+</div>
+
+</div>
+</div>
+
 </div>
     {% endraw %}
 
@@ -2494,7 +2573,7 @@ <h3 id="WHERE">WHERE<a class="anchor-link" href="#WHERE"> </a></h3>
 
 
 <div class="output_markdown rendered_html output_subarea ">
-<h4 id="format_where" class="doc_header"><code>format_where</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L320" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>format_where</code>(<strong><code>s</code></strong>)</p>
+<h4 id="format_where" class="doc_header"><code>format_where</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L323" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>format_where</code>(<strong><code>s</code></strong>)</p>
 </blockquote>
 <p>Format WHERE statement line <code>s</code></p>
 
@@ -2569,7 +2648,7 @@ <h2 id="Format-all-statements">Format all statements<a class="anchor-link" href=
 
 
 <div class="output_markdown rendered_html output_subarea ">
-<h4 id="format_statement_line" class="doc_header"><code>format_statement_line</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L330" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>format_statement_line</code>(<strong><code>s</code></strong>)</p>
+<h4 id="format_statement_line" class="doc_header"><code>format_statement_line</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L333" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>format_statement_line</code>(<strong><code>s</code></strong>)</p>
 </blockquote>
 <p>Format statement line <code>s</code></p>
 
@@ -2713,7 +2792,7 @@ <h4 id="format_statement_line" class="doc_header"><code>format_statement_line</c
 
 
 <div class="output_markdown rendered_html output_subarea ">
-<h4 id="format_statements" class="doc_header"><code>format_statements</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L348" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>format_statements</code>(<strong><code>s</code></strong>)</p>
+<h4 id="format_statements" class="doc_header"><code>format_statements</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L351" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>format_statements</code>(<strong><code>s</code></strong>)</p>
 </blockquote>
 <p>Format statements lines <code>s</code></p>
 
@@ -2791,7 +2870,7 @@ <h3 id="Write-a-;-at-the-end-of-query">Write a ; at the end of query<a class="an
 
 
 <div class="output_markdown rendered_html output_subarea ">
-<h4 id="add_ending_semicolon" class="doc_header"><code>add_ending_semicolon</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L358" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>add_ending_semicolon</code>(<strong><code>s</code></strong>)</p>
+<h4 id="add_ending_semicolon" class="doc_header"><code>add_ending_semicolon</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L361" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>add_ending_semicolon</code>(<strong><code>s</code></strong>)</p>
 </blockquote>
 <p>Add ending semicolon for SQL query <code>s</code></p>
 
@@ -2957,7 +3036,7 @@ <h2 id="Putting-everything-together">Putting everything together<a class="anchor
 
 
 <div class="output_markdown rendered_html output_subarea ">
-<h4 id="format_simple_sql" class="doc_header"><code>format_simple_sql</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L366" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>format_simple_sql</code>(<strong><code>s</code></strong>, <strong><code>add_semicolon</code></strong>=<em><code>True</code></em>)</p>
+<h4 id="format_simple_sql" class="doc_header"><code>format_simple_sql</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L369" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>format_simple_sql</code>(<strong><code>s</code></strong>, <strong><code>add_semicolon</code></strong>=<em><code>True</code></em>)</p>
 </blockquote>
 <p>Format a simple SQL query without subqueries <code>s</code>.
 If <code>add_semicolon</code> is True, then add a semicolon at the end</p>
@@ -3018,7 +3097,8 @@ <h4 id="format_simple_sql" class="doc_header"><code>format_simple_sql</code><a h
     LEFT JOIN table2 as b -- and here a comment
         ON a.asdf = b.asdf /* joiN this way */
     INNER JOIN table3 as c
-        ON a.asdf = c.asdf
+        ON a.asdf = c.asdf and
+           a.qwer = b.qwer
 WHERE  a.asdf = 1 -- comment this
    and b.qwer = 2
    and a.asdf &lt;= 1 --comment that
@@ -3180,7 +3260,7 @@ <h3 id="Main-function-handling-queries-with-subqueries">Main function handling q
 
 
 <div class="output_markdown rendered_html output_subarea ">
-<h4 id="format_sql" class="doc_header"><code>format_sql</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L380" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>format_sql</code>(<strong><code>s</code></strong>, <strong><code>add_semicolon</code></strong>=<em><code>True</code></em>)</p>
+<h4 id="format_sql" class="doc_header"><code>format_sql</code><a href="https://github.com/PabloRMira/sql_formatter/tree/master/sql_formatter/core.py#L384" class="source_link" style="float:right">[source]</a></h4><blockquote><p><code>format_sql</code>(<strong><code>s</code></strong>, <strong><code>add_semicolon</code></strong>=<em><code>True</code></em>)</p>
 </blockquote>
 <p>Format SQL query with subqueries. If <code>add_semicolon</code> is True then add a semicolon at the end</p>
 
@@ -3293,7 +3373,8 @@ <h4 id="format_sql" class="doc_header"><code>format_sql</code><a href="https://g
     LEFT JOIN table2 as b -- and here a comment
         ON a.asdf = b.asdf /* joiN this way */
     INNER JOIN table3 as c
-        ON a.asdf = c.asdf
+        ON a.asdf = c.asdf and
+           a.qwer = b.qwer
 WHERE  a.asdf = 1 -- comment this
    and b.qwer = 2
    and a.asdf &lt;= 1 --comment that

diff --git a/nbs/00_core.ipynb b/nbs/00_core.ipynb
@@ -4,16 +4,7 @@
    "cell_type": "code",
    "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "The autoreload extension is already loaded. To reload it, use:\n",
-      "  %reload_ext autoreload\n"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "#hide\n",
     "%load_ext autoreload\n",
@@ -96,7 +87,7 @@
     "table2 as b -- and here a comment\n",
     "    on a.asdf = b.asdf  /* joiN this way */\n",
     "    inner join table3 as c\n",
-    "on a.asdf=c.asdf\n",
+    "on a.asdf=c.asdf and a.qwer= b.qwer\n",
     "whEre a.asdf= 1 -- comment this\n",
     "anD b.qwer =2 and a.asdf<=1 --comment that\n",
     "or b.qwer>=5\n",
@@ -133,7 +124,8 @@
     "    LEFT JOIN table2 as b -- and here a comment\n",
     "        ON a.asdf = b.asdf /* joiN this way */\n",
     "    INNER JOIN table3 as c\n",
-    "        ON a.asdf = c.asdf\n",
+    "        ON a.asdf = c.asdf and\n",
+    "           a.qwer = b.qwer\n",
     "WHERE  a.asdf = 1 -- comment this\n",
     "   and b.qwer = 2\n",
     "   and a.asdf <= 1 --comment that\n",
@@ -1951,8 +1943,11 @@
     "#export\n",
     "def format_on(s):\n",
     "    \"Format ON statement line `s`\"\n",
-    "    s = add_whitespaces_between_symbols(s)  # add whitespaces between symbols in join    \n",
+    "    s = add_whitespaces_between_symbols(s)  # add whitespaces between symbols in join\n",
+    "    s = s.replace(\"[C]\", \" \")\n",
     "    s = \"        \" + s  # add indentation\n",
+    "    s = re.sub(r\"and\\s\", r\"and\\n\" + \" \" * 11, s, flags=re.I)  # add new line before every 'and' and indentation\n",
+    "    s = re.sub(r\"or\\s\", r\"or\\n\" + \" \" * 11, s, flags=re.I)  # add new line before every 'or' and indentation\n",
     "    return s"
    ]
   },
@@ -1973,6 +1968,56 @@
     "assert_and_print(format_on(\"on a.asdf =b.asdf\"), \"        on a.asdf = b.asdf\")"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "        on a.asdf = b.asdf and\n",
+      "           a.qwer = b.qwer\n"
+     ]
+    }
+   ],
+   "source": [
+    "assert_and_print(\n",
+    "    format_on(\"on a.asdf =b.asdf and a.qwer=b.qwer\"), \n",
+    "\"\"\"\n",
+    "        on a.asdf = b.asdf and\n",
+    "           a.qwer = b.qwer\n",
+    "\"\"\".strip(\"\\n\")\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "        on a.asdf = b.asdf and\n",
+      "           a.qwer = b.qwer or\n",
+      "           a.qwer2 = b.qwer2\n"
+     ]
+    }
+   ],
+   "source": [
+    "assert_and_print(\n",
+    "    format_on(\"on a.asdf =b.asdf and a.qwer=b.qwer or a.qwer2= b.qwer2\"), \n",
+    "\"\"\"\n",
+    "        on a.asdf = b.asdf and\n",
+    "           a.qwer = b.qwer or\n",
+    "           a.qwer2 = b.qwer2\n",
+    "\"\"\".strip(\"\\n\")\n",
+    ")"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -2296,7 +2341,8 @@
     "    s = preformat_statements(s)  # add breaklines for the main statements\n",
     "    s = format_statements(s)  # format statements\n",
     "    s = re.sub(r\"\\[C\\]\", \"\", s)  # replace remaining [C]\n",
-    "    s = re.sub(r\"\\[CS\\]\", \"\\n\", s)  # replace remainig [CS]\n",
+    "    s = re.sub(r\"\\[CS\\]\", \"\\n\", s)  # replace remaining [CS]\n",
+    "    s = re.sub(r\"\\s+\\n\", \"\\n\", s)  # replace redundant whitespaces before newline\n",
     "    if add_semicolon:\n",
     "        s = add_ending_semicolon(s)  # add ending semicolon if not there yet\n",
     "    return s"
@@ -2327,7 +2373,8 @@
       "    LEFT JOIN table2 as b -- and here a comment\n",
       "        ON a.asdf = b.asdf /* joiN this way */\n",
       "    INNER JOIN table3 as c\n",
-      "        ON a.asdf = c.asdf\n",
+      "        ON a.asdf = c.asdf and\n",
+      "           a.qwer = b.qwer\n",
       "WHERE  a.asdf = 1 -- comment this\n",
       "   and b.qwer = 2\n",
       "   and a.asdf <= 1 --comment that\n",
@@ -2545,7 +2592,8 @@
       "    LEFT JOIN table2 as b -- and here a comment\n",
       "        ON a.asdf = b.asdf /* joiN this way */\n",
       "    INNER JOIN table3 as c\n",
-      "        ON a.asdf = c.asdf\n",
+      "        ON a.asdf = c.asdf and\n",
+      "           a.qwer = b.qwer\n",
       "WHERE  a.asdf = 1 -- comment this\n",
       "   and b.qwer = 2\n",
       "   and a.asdf <= 1 --comment that\n",

diff --git a/sql_formatter/core.py b/sql_formatter/core.py
@@ -313,7 +313,10 @@ def format_join(s):
 def format_on(s):
     "Format ON statement line `s`"
     s = add_whitespaces_between_symbols(s)  # add whitespaces between symbols in join
+    s = s.replace("[C]", " ")
     s = "        " + s  # add indentation
+    s = re.sub(r"and\s", r"and\n" + " " * 11, s, flags=re.I)  # add new line before every 'and' and indentation
+    s = re.sub(r"or\s", r"or\n" + " " * 11, s, flags=re.I)  # add new line before every 'or' and indentation
     return s
 
 # Cell
@@ -371,7 +374,8 @@ def format_simple_sql(s, add_semicolon=True):
     s = preformat_statements(s)  # add breaklines for the main statements
     s = format_statements(s)  # format statements
     s = re.sub(r"\[C\]", "", s)  # replace remaining [C]
-    s = re.sub(r"\[CS\]", "\n", s)  # replace remainig [CS]
+    s = re.sub(r"\[CS\]", "\n", s)  # replace remaining [CS]
+    s = re.sub(r"\s+\n", "\n", s)  # replace redundant whitespaces before newline
     if add_semicolon:
         s = add_ending_semicolon(s)  # add ending semicolon if not there yet
     return s