<entry id="2300485" stat="A" corpus="jmdict" type="jmdict">
<ent_corp type="jmdict">jmdict</ent_corp>
<ent_seq>1049180</ent_seq>
<k_ele>
<keb>珈琲</keb>
<ke_inf>&ateji;</ke_inf>
</k_ele>
<r_ele>
<reb>コーヒー</reb>
<re_pri>gai1</re_pri>
<re_pri>ichi1</re_pri>
</r_ele>
<sense>
<pos>&n;</pos>
<misc>&uk;</misc>
<lsource xml:lang="eng">coffee</lsource>
<lsource xml:lang="dut">koffie</lsource>
<gloss>coffee</gloss>
</sense>
<info>
<audit time="2017-04-21 09:02:22" stat="A" unap="true">
<upd_uid>johan</upd_uid>
<upd_name>Johan Råde</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_detl>I think we can view コーヒー as a kana form of 珈琲</upd_detl>
<upd_diff>@@ -14 +13,0 @@
-&lt;re_nokanji/&gt;</upd_diff>
</audit>
<audit time="2017-04-25 23:41:54" stat="A" unap="true">
<upd_uid>jwb</upd_uid>
<upd_name>Jim Breen</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_detl>In that case the rather artificial こーひー should go.</upd_detl>
<upd_diff>@@ -9,2 +8,0 @@
-&lt;ke_pri&gt;news2&lt;/ke_pri&gt;
-&lt;ke_pri&gt;nf35&lt;/ke_pri&gt;
@@ -16,5 +13,0 @@
-&lt;/r_ele&gt;
-&lt;r_ele&gt;
-&lt;reb&gt;こーひー&lt;/reb&gt;
-&lt;re_pri&gt;news2&lt;/re_pri&gt;
-&lt;re_pri&gt;nf35&lt;/re_pri&gt;</upd_diff>
</audit>
<audit time="2017-04-26 04:36:14" stat="A" unap="true">
<upd_name>Johan Råde</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_detl>0.2%, but still fairly common in absolute numbers</upd_detl>
<upd_refs>googits
 コーヒー  227 M
 こーひー  468 K</upd_refs>
<upd_diff>@@ -14,0 +15,3 @@
+&lt;r_ele&gt;
+&lt;reb&gt;こーひー&lt;/reb&gt;
+&lt;/r_ele&gt;</upd_diff>
</audit>
<audit time="2017-04-26 11:51:57" stat="A">
<upd_uid>jwb</upd_uid>
<upd_name>Jim Breen</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_refs>117 unique Googits, not that they mean anything these days.</upd_refs>
</audit>
<audit time="2017-04-26 12:08:32" stat="A" unap="true">
<upd_name>Johan Råde</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_detl>Jim, How did you get the number 117?
What search string did you use?</upd_detl>
</audit>
<audit time="2017-04-26 18:35:45" stat="A" unap="true">
<upd_name>Robin Scott</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_detl>Presumably by clicking through to the last page of results. (For me it drops from 473k to 121).

But this is the number of results Google will show you, not the total number of matches.
Google never retrieves more than a few hundreds results (I get 239 for "cat").

The difficulty is in knowing which number to trust. The hit count shown on the first page is only an estimate and often wildly inaccurate.
In the case of "cat", 2.3 billion is obviously much closer to the real number than 239, but try an exact match search on something really obscure.
I searched for the clearly ungrammatical "りんごを落ちる" and got "38,300 results".
But click on page 2 and it ends there. 15 results.

Both these numbers are probably wrong, but it's anyone's guess as to how wrong.

Something to keep in mind when looking at Google hits.</upd_detl>
</audit>
<audit time="2017-04-26 20:48:02" stat="A" unap="true">
<upd_uid>johan</upd_uid>
<upd_name>Johan Råde</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_detl>let's skip こーひー then</upd_detl>
<upd_diff>@@ -15,3 +14,0 @@
-&lt;r_ele&gt;
-&lt;reb&gt;こーひー&lt;/reb&gt;
-&lt;/r_ele&gt;</upd_diff>
</audit>
<audit time="2017-04-30 06:44:57" stat="A">
<upd_uid>jwb</upd_uid>
<upd_name>Jim Breen</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_detl>OK.</upd_detl>
</audit>
<audit time="2024-05-09 00:03:19" stat="A">
<upd_uid>Marcus</upd_uid>
<upd_name>Marcus Richert</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_diff>@@ -17 +16,0 @@
-&lt;pos&gt;&amp;adj-no;&lt;/pos&gt;</upd_diff>
</audit>
<audit time="2024-05-12 21:54:31" stat="A" unap="true">
<upd_uid>stephen</upd_uid>
<upd_name>Stephen Kraus</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_detl>Didn't receive an answer to my question on ありがとう (1586820). If we're going to drop the priority tags on 有難う, seems like they should be dropped on 珈琲 as well.</upd_detl>
<upd_refs>Google N-gram Corpus Counts
╭─ーーーーー─┬────────────┬───────╮
│ 珈琲　　　 │  2,390,900 │ 15.4% │
│ コーヒー　 │ 13,171,397 │ 84.6% │
├─ーーーーー─┼────────────┼───────┤
│ 有難う　　 │  8,616,052 │  8.1% │
│ 有り難う　 │  2,274,621 │  2.1% │
│ 有りがとう │      8,914 │  0.0% │
│ ありがとう │ 95,539,750 │ 89.8% │
╰─ーーーーー─┴────────────┴───────╯</upd_refs>
<upd_diff>@@ -7,2 +6,0 @@
-&lt;ke_pri&gt;gai1&lt;/ke_pri&gt;
-&lt;ke_pri&gt;ichi1&lt;/ke_pri&gt;</upd_diff>
</audit>
<audit time="2024-05-12 21:58:13" stat="A">
<upd_uid>jwb</upd_uid>
<upd_name>Jim Breen</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_detl>Sorry if my response was unclear. I think it's fine to drop the tags from the kanji fields in these uk entries.</upd_detl>
</audit>
<audit time="2024-05-12 23:34:13" stat="A">
<upd_uid>stephen</upd_uid>
<upd_name>Stephen Kraus</upd_name>
<upd_email>...address hidden...</upd_email>
<upd_detl>Marcus recently amended both entries and dropped the tags in ありがとう but not コーヒー, so I was just wondering if there was a reason for that difference.

Sorry if I sound annoyed; I need to update my dictionary build scripts to handle this sort of setup, and I don't understand the point of it all if the same information is already provided by the [uk] tag. But it sounds like that ship has sailed, so I'll stop griping.</upd_detl>
</audit>
</info>
</entry>