This is an automated email from the ASF dual-hosted git repository.
github-bot pushed a commit to branch gh-pages
in repository https://gitbox.apache.org/repos/asf/iceberg-rust.git
The following commit(s) were added to refs/heads/gh-pages by this push:
new 783923f9a deploy: 5cf85dab903e171a077481597be1e49a1159715a
783923f9a is described below
commit 783923f9a47ac8837f1107149d74653d321a9f6a
Author: liurenjie1024 <[email protected]>
AuthorDate: Wed Jan 14 00:03:13 2026 +0000
deploy: 5cf85dab903e171a077481597be1e49a1159715a
---
api/src/iceberg_datafusion/schema.rs.html | 460 ++++++++++++++++-----
.../iceberg_sqllogictest/engine/datafusion.rs.html | 99 ++---
2 files changed, 385 insertions(+), 174 deletions(-)
diff --git a/api/src/iceberg_datafusion/schema.rs.html
b/api/src/iceberg_datafusion/schema.rs.html
index 8ce943366..794963741 100644
--- a/api/src/iceberg_datafusion/schema.rs.html
+++ b/api/src/iceberg_datafusion/schema.rs.html
@@ -16,118 +16,356 @@
<a href=#16 id=16 data-nosnippet>16</a>// under the License.
<a href=#17 id=17 data-nosnippet>17</a>
<a href=#18 id=18 data-nosnippet>18</a></span><span class="kw">use
</span>std::any::Any;
-<a href=#19 id=19 data-nosnippet>19</a><span class="kw">use
</span>std::collections::HashMap;
-<a href=#20 id=20 data-nosnippet>20</a><span class="kw">use
</span>std::sync::Arc;
-<a href=#21 id=21 data-nosnippet>21</a>
-<a href=#22 id=22 data-nosnippet>22</a><span class="kw">use
</span>async_trait::async_trait;
+<a href=#19 id=19 data-nosnippet>19</a><span class="kw">use
</span>std::sync::Arc;
+<a href=#20 id=20 data-nosnippet>20</a>
+<a href=#21 id=21 data-nosnippet>21</a><span class="kw">use
</span>async_trait::async_trait;
+<a href=#22 id=22 data-nosnippet>22</a><span class="kw">use
</span>dashmap::DashMap;
<a href=#23 id=23 data-nosnippet>23</a><span class="kw">use
</span>datafusion::catalog::SchemaProvider;
<a href=#24 id=24 data-nosnippet>24</a><span class="kw">use
</span>datafusion::datasource::TableProvider;
<a href=#25 id=25 data-nosnippet>25</a><span class="kw">use
</span>datafusion::error::{DataFusionError, <span
class="prelude-ty">Result</span> <span class="kw">as </span>DFResult};
-<a href=#26 id=26 data-nosnippet>26</a><span class="kw">use
</span>futures::future::try_join_all;
-<a href=#27 id=27 data-nosnippet>27</a><span class="kw">use
</span>iceberg::inspect::MetadataTableType;
-<a href=#28 id=28 data-nosnippet>28</a><span class="kw">use
</span>iceberg::{Catalog, NamespaceIdent, <span
class="prelude-ty">Result</span>};
-<a href=#29 id=29 data-nosnippet>29</a>
-<a href=#30 id=30 data-nosnippet>30</a><span class="kw">use </span><span
class="kw">crate</span>::table::IcebergTableProvider;
-<a href=#31 id=31 data-nosnippet>31</a><span class="kw">use </span><span
class="kw">crate</span>::to_datafusion_error;
-<a href=#32 id=32 data-nosnippet>32</a>
-<a href=#33 id=33 data-nosnippet>33</a><span class="doccomment">/// Represents
a [`SchemaProvider`] for the Iceberg [`Catalog`], managing
-<a href=#34 id=34 data-nosnippet>34</a>/// access to table providers within a
specific namespace.
-<a href=#35 id=35 data-nosnippet>35</a></span><span
class="attr">#[derive(Debug)]
-<a href=#36 id=36 data-nosnippet>36</a></span><span
class="kw">pub</span>(<span class="kw">crate</span>) <span class="kw">struct
</span>IcebergSchemaProvider {
-<a href=#37 id=37 data-nosnippet>37</a> <span class="doccomment">/// A
`HashMap` where keys are table names
-<a href=#38 id=38 data-nosnippet>38</a> /// and values are dynamic
references to objects implementing the
-<a href=#39 id=39 data-nosnippet>39</a> /// [`TableProvider`] trait.
-<a href=#40 id=40 data-nosnippet>40</a> </span>tables: HashMap<String,
Arc<IcebergTableProvider>>,
-<a href=#41 id=41 data-nosnippet>41</a>}
-<a href=#42 id=42 data-nosnippet>42</a>
-<a href=#43 id=43 data-nosnippet>43</a><span class="kw">impl
</span>IcebergSchemaProvider {
-<a href=#44 id=44 data-nosnippet>44</a> <span class="doccomment">///
Asynchronously tries to construct a new [`IcebergSchemaProvider`]
-<a href=#45 id=45 data-nosnippet>45</a> /// using the given client to fetch
and initialize table providers for
-<a href=#46 id=46 data-nosnippet>46</a> /// the provided namespace in the
Iceberg [`Catalog`].
-<a href=#47 id=47 data-nosnippet>47</a> ///
-<a href=#48 id=48 data-nosnippet>48</a> /// This method retrieves a list of
table names
-<a href=#49 id=49 data-nosnippet>49</a> /// attempts to create a table
provider for each table name, and
-<a href=#50 id=50 data-nosnippet>50</a> /// collects these providers into a
`HashMap`.
-<a href=#51 id=51 data-nosnippet>51</a> </span><span
class="kw">pub</span>(<span class="kw">crate</span>) <span class="kw">async fn
</span>try_new(
-<a href=#52 id=52 data-nosnippet>52</a> client: Arc<<span
class="kw">dyn </span>Catalog>,
-<a href=#53 id=53 data-nosnippet>53</a> namespace: NamespaceIdent,
-<a href=#54 id=54 data-nosnippet>54</a> ) -> <span
class="prelude-ty">Result</span><<span class="self">Self</span>> {
-<a href=#55 id=55 data-nosnippet>55</a> <span class="comment">// TODO:
-<a href=#56 id=56 data-nosnippet>56</a> // Tables and providers should
be cached based on table_name
-<a href=#57 id=57 data-nosnippet>57</a> // if we have a cache miss; we
update our internal cache & check again
-<a href=#58 id=58 data-nosnippet>58</a> // As of right now; tables
might become stale.
-<a href=#59 id=59 data-nosnippet>59</a> </span><span class="kw">let
</span>table_names: Vec<<span class="kw">_</span>> = client
-<a href=#60 id=60 data-nosnippet>60</a> .list_tables(<span
class="kw-2">&</span>namespace)
-<a href=#61 id=61 data-nosnippet>61</a> .<span
class="kw">await</span><span class="question-mark">?
-<a href=#62 id=62 data-nosnippet>62</a> </span>.iter()
-<a href=#63 id=63 data-nosnippet>63</a> .map(|tbl|
tbl.name().to_string())
-<a href=#64 id=64 data-nosnippet>64</a> .collect();
-<a href=#65 id=65 data-nosnippet>65</a>
-<a href=#66 id=66 data-nosnippet>66</a> <span class="kw">let
</span>providers = try_join_all(
-<a href=#67 id=67 data-nosnippet>67</a> table_names
-<a href=#68 id=68 data-nosnippet>68</a> .iter()
-<a href=#69 id=69 data-nosnippet>69</a> .map(|name|
IcebergTableProvider::try_new(client.clone(), namespace.clone(), name))
-<a href=#70 id=70 data-nosnippet>70</a>
.collect::<Vec<<span class="kw">_</span>>>(),
-<a href=#71 id=71 data-nosnippet>71</a> )
-<a href=#72 id=72 data-nosnippet>72</a> .<span
class="kw">await</span><span class="question-mark">?</span>;
-<a href=#73 id=73 data-nosnippet>73</a>
-<a href=#74 id=74 data-nosnippet>74</a> <span class="kw">let
</span>tables: HashMap<String, Arc<IcebergTableProvider>> =
table_names
-<a href=#75 id=75 data-nosnippet>75</a> .into_iter()
-<a href=#76 id=76 data-nosnippet>76</a> .zip(providers.into_iter())
-<a href=#77 id=77 data-nosnippet>77</a> .map(|(name, provider)|
(name, Arc::new(provider)))
-<a href=#78 id=78 data-nosnippet>78</a> .collect();
-<a href=#79 id=79 data-nosnippet>79</a>
-<a href=#80 id=80 data-nosnippet>80</a> <span
class="prelude-val">Ok</span>(IcebergSchemaProvider { tables })
-<a href=#81 id=81 data-nosnippet>81</a> }
-<a href=#82 id=82 data-nosnippet>82</a>}
-<a href=#83 id=83 data-nosnippet>83</a>
-<a href=#84 id=84 data-nosnippet>84</a><span class="attr">#[async_trait]
-<a href=#85 id=85 data-nosnippet>85</a></span><span class="kw">impl
</span>SchemaProvider <span class="kw">for </span>IcebergSchemaProvider {
-<a href=#86 id=86 data-nosnippet>86</a> <span class="kw">fn
</span>as_any(<span class="kw-2">&</span><span class="self">self</span>)
-> <span class="kw-2">&</span><span class="kw">dyn </span>Any {
-<a href=#87 id=87 data-nosnippet>87</a> <span class="self">self
-<a href=#88 id=88 data-nosnippet>88</a> </span>}
-<a href=#89 id=89 data-nosnippet>89</a>
-<a href=#90 id=90 data-nosnippet>90</a> <span class="kw">fn
</span>table_names(<span class="kw-2">&</span><span
class="self">self</span>) -> Vec<String> {
-<a href=#91 id=91 data-nosnippet>91</a> <span
class="self">self</span>.tables
-<a href=#92 id=92 data-nosnippet>92</a> .keys()
-<a href=#93 id=93 data-nosnippet>93</a> .flat_map(|table_name| {
-<a href=#94 id=94 data-nosnippet>94</a> [table_name.clone()]
-<a href=#95 id=95 data-nosnippet>95</a> .into_iter()
-<a href=#96 id=96 data-nosnippet>96</a>
.chain(MetadataTableType::all_types().map(|metadata_table_name| {
-<a href=#97 id=97 data-nosnippet>97</a> <span
class="macro">format!</span>(<span class="string">"{}${}"</span>,
table_name.clone(), metadata_table_name.as_str())
-<a href=#98 id=98 data-nosnippet>98</a> }))
-<a href=#99 id=99 data-nosnippet>99</a> })
-<a href=#100 id=100 data-nosnippet>100</a> .collect()
-<a href=#101 id=101 data-nosnippet>101</a> }
-<a href=#102 id=102 data-nosnippet>102</a>
-<a href=#103 id=103 data-nosnippet>103</a> <span class="kw">fn
</span>table_exist(<span class="kw-2">&</span><span
class="self">self</span>, name: <span class="kw-2">&</span>str) -> bool {
-<a href=#104 id=104 data-nosnippet>104</a> <span class="kw">if let
</span><span class="prelude-val">Some</span>((table_name, metadata_table_name))
= name.split_once(<span class="string">'$'</span>) {
-<a href=#105 id=105 data-nosnippet>105</a> <span
class="self">self</span>.tables.contains_key(table_name)
-<a href=#106 id=106 data-nosnippet>106</a> &&
MetadataTableType::try_from(metadata_table_name).is_ok()
-<a href=#107 id=107 data-nosnippet>107</a> } <span class="kw">else
</span>{
-<a href=#108 id=108 data-nosnippet>108</a> <span
class="self">self</span>.tables.contains_key(name)
-<a href=#109 id=109 data-nosnippet>109</a> }
-<a href=#110 id=110 data-nosnippet>110</a> }
-<a href=#111 id=111 data-nosnippet>111</a>
-<a href=#112 id=112 data-nosnippet>112</a> <span class="kw">async fn
</span>table(<span class="kw-2">&</span><span class="self">self</span>,
name: <span class="kw-2">&</span>str) -> DFResult<<span
class="prelude-ty">Option</span><Arc<<span class="kw">dyn
</span>TableProvider>>> {
-<a href=#113 id=113 data-nosnippet>113</a> <span class="kw">if let
</span><span class="prelude-val">Some</span>((table_name, metadata_table_name))
= name.split_once(<span class="string">'$'</span>) {
-<a href=#114 id=114 data-nosnippet>114</a> <span class="kw">let
</span>metadata_table_type =
-<a href=#115 id=115 data-nosnippet>115</a>
MetadataTableType::try_from(metadata_table_name).map_err(DataFusionError::Plan)<span
class="question-mark">?</span>;
-<a href=#116 id=116 data-nosnippet>116</a> <span class="kw">if let
</span><span class="prelude-val">Some</span>(table) = <span
class="self">self</span>.tables.get(table_name) {
-<a href=#117 id=117 data-nosnippet>117</a> <span class="kw">let
</span>metadata_table = table
-<a href=#118 id=118 data-nosnippet>118</a>
.metadata_table(metadata_table_type)
-<a href=#119 id=119 data-nosnippet>119</a> .<span
class="kw">await
-<a href=#120 id=120 data-nosnippet>120</a>
</span>.map_err(to_datafusion_error)<span class="question-mark">?</span>;
-<a href=#121 id=121 data-nosnippet>121</a> <span
class="kw">return </span><span class="prelude-val">Ok</span>(<span
class="prelude-val">Some</span>(Arc::new(metadata_table)));
-<a href=#122 id=122 data-nosnippet>122</a> } <span class="kw">else
</span>{
-<a href=#123 id=123 data-nosnippet>123</a> <span
class="kw">return </span><span class="prelude-val">Ok</span>(<span
class="prelude-val">None</span>);
-<a href=#124 id=124 data-nosnippet>124</a> }
-<a href=#125 id=125 data-nosnippet>125</a> }
+<a href=#26 id=26 data-nosnippet>26</a><span class="kw">use
</span>datafusion::execution::TaskContext;
+<a href=#27 id=27 data-nosnippet>27</a><span class="kw">use
</span>datafusion::prelude::SessionContext;
+<a href=#28 id=28 data-nosnippet>28</a><span class="kw">use
</span>futures::StreamExt;
+<a href=#29 id=29 data-nosnippet>29</a><span class="kw">use
</span>futures::future::try_join_all;
+<a href=#30 id=30 data-nosnippet>30</a><span class="kw">use
</span>iceberg::arrow::arrow_schema_to_schema_auto_assign_ids;
+<a href=#31 id=31 data-nosnippet>31</a><span class="kw">use
</span>iceberg::inspect::MetadataTableType;
+<a href=#32 id=32 data-nosnippet>32</a><span class="kw">use
</span>iceberg::{Catalog, Error, ErrorKind, NamespaceIdent, <span
class="prelude-ty">Result</span>, TableCreation};
+<a href=#33 id=33 data-nosnippet>33</a>
+<a href=#34 id=34 data-nosnippet>34</a><span class="kw">use </span><span
class="kw">crate</span>::table::IcebergTableProvider;
+<a href=#35 id=35 data-nosnippet>35</a><span class="kw">use </span><span
class="kw">crate</span>::to_datafusion_error;
+<a href=#36 id=36 data-nosnippet>36</a>
+<a href=#37 id=37 data-nosnippet>37</a><span class="doccomment">/// Represents
a [`SchemaProvider`] for the Iceberg [`Catalog`], managing
+<a href=#38 id=38 data-nosnippet>38</a>/// access to table providers within a
specific namespace.
+<a href=#39 id=39 data-nosnippet>39</a></span><span
class="attr">#[derive(Debug)]
+<a href=#40 id=40 data-nosnippet>40</a></span><span
class="kw">pub</span>(<span class="kw">crate</span>) <span class="kw">struct
</span>IcebergSchemaProvider {
+<a href=#41 id=41 data-nosnippet>41</a> <span class="doccomment">///
Reference to the Iceberg catalog
+<a href=#42 id=42 data-nosnippet>42</a> </span>catalog: Arc<<span
class="kw">dyn </span>Catalog>,
+<a href=#43 id=43 data-nosnippet>43</a> <span class="doccomment">/// The
namespace this schema represents
+<a href=#44 id=44 data-nosnippet>44</a> </span>namespace: NamespaceIdent,
+<a href=#45 id=45 data-nosnippet>45</a> <span class="doccomment">/// A
concurrent map where keys are table names
+<a href=#46 id=46 data-nosnippet>46</a> /// and values are dynamic
references to objects implementing the
+<a href=#47 id=47 data-nosnippet>47</a> /// [`TableProvider`] trait.
+<a href=#48 id=48 data-nosnippet>48</a> /// Wrapped in Arc to allow sharing
across async boundaries in register_table.
+<a href=#49 id=49 data-nosnippet>49</a> </span>tables:
Arc<DashMap<String, Arc<IcebergTableProvider>>>,
+<a href=#50 id=50 data-nosnippet>50</a>}
+<a href=#51 id=51 data-nosnippet>51</a>
+<a href=#52 id=52 data-nosnippet>52</a><span class="kw">impl
</span>IcebergSchemaProvider {
+<a href=#53 id=53 data-nosnippet>53</a> <span class="doccomment">///
Asynchronously tries to construct a new [`IcebergSchemaProvider`]
+<a href=#54 id=54 data-nosnippet>54</a> /// using the given client to fetch
and initialize table providers for
+<a href=#55 id=55 data-nosnippet>55</a> /// the provided namespace in the
Iceberg [`Catalog`].
+<a href=#56 id=56 data-nosnippet>56</a> ///
+<a href=#57 id=57 data-nosnippet>57</a> /// This method retrieves a list of
table names
+<a href=#58 id=58 data-nosnippet>58</a> /// attempts to create a table
provider for each table name, and
+<a href=#59 id=59 data-nosnippet>59</a> /// collects these providers into a
`HashMap`.
+<a href=#60 id=60 data-nosnippet>60</a> </span><span
class="kw">pub</span>(<span class="kw">crate</span>) <span class="kw">async fn
</span>try_new(
+<a href=#61 id=61 data-nosnippet>61</a> client: Arc<<span
class="kw">dyn </span>Catalog>,
+<a href=#62 id=62 data-nosnippet>62</a> namespace: NamespaceIdent,
+<a href=#63 id=63 data-nosnippet>63</a> ) -> <span
class="prelude-ty">Result</span><<span class="self">Self</span>> {
+<a href=#64 id=64 data-nosnippet>64</a> <span class="comment">// TODO:
+<a href=#65 id=65 data-nosnippet>65</a> // Tables and providers should
be cached based on table_name
+<a href=#66 id=66 data-nosnippet>66</a> // if we have a cache miss; we
update our internal cache & check again
+<a href=#67 id=67 data-nosnippet>67</a> // As of right now; tables
might become stale.
+<a href=#68 id=68 data-nosnippet>68</a> </span><span class="kw">let
</span>table_names: Vec<<span class="kw">_</span>> = client
+<a href=#69 id=69 data-nosnippet>69</a> .list_tables(<span
class="kw-2">&</span>namespace)
+<a href=#70 id=70 data-nosnippet>70</a> .<span
class="kw">await</span><span class="question-mark">?
+<a href=#71 id=71 data-nosnippet>71</a> </span>.iter()
+<a href=#72 id=72 data-nosnippet>72</a> .map(|tbl|
tbl.name().to_string())
+<a href=#73 id=73 data-nosnippet>73</a> .collect();
+<a href=#74 id=74 data-nosnippet>74</a>
+<a href=#75 id=75 data-nosnippet>75</a> <span class="kw">let
</span>providers = try_join_all(
+<a href=#76 id=76 data-nosnippet>76</a> table_names
+<a href=#77 id=77 data-nosnippet>77</a> .iter()
+<a href=#78 id=78 data-nosnippet>78</a> .map(|name|
IcebergTableProvider::try_new(client.clone(), namespace.clone(), name))
+<a href=#79 id=79 data-nosnippet>79</a>
.collect::<Vec<<span class="kw">_</span>>>(),
+<a href=#80 id=80 data-nosnippet>80</a> )
+<a href=#81 id=81 data-nosnippet>81</a> .<span
class="kw">await</span><span class="question-mark">?</span>;
+<a href=#82 id=82 data-nosnippet>82</a>
+<a href=#83 id=83 data-nosnippet>83</a> <span class="kw">let
</span>tables = Arc::new(DashMap::new());
+<a href=#84 id=84 data-nosnippet>84</a> <span class="kw">for
</span>(name, provider) <span class="kw">in
</span>table_names.into_iter().zip(providers.into_iter()) {
+<a href=#85 id=85 data-nosnippet>85</a> tables.insert(name,
Arc::new(provider));
+<a href=#86 id=86 data-nosnippet>86</a> }
+<a href=#87 id=87 data-nosnippet>87</a>
+<a href=#88 id=88 data-nosnippet>88</a> <span
class="prelude-val">Ok</span>(IcebergSchemaProvider {
+<a href=#89 id=89 data-nosnippet>89</a> catalog: client,
+<a href=#90 id=90 data-nosnippet>90</a> namespace,
+<a href=#91 id=91 data-nosnippet>91</a> tables,
+<a href=#92 id=92 data-nosnippet>92</a> })
+<a href=#93 id=93 data-nosnippet>93</a> }
+<a href=#94 id=94 data-nosnippet>94</a>}
+<a href=#95 id=95 data-nosnippet>95</a>
+<a href=#96 id=96 data-nosnippet>96</a><span class="attr">#[async_trait]
+<a href=#97 id=97 data-nosnippet>97</a></span><span class="kw">impl
</span>SchemaProvider <span class="kw">for </span>IcebergSchemaProvider {
+<a href=#98 id=98 data-nosnippet>98</a> <span class="kw">fn
</span>as_any(<span class="kw-2">&</span><span class="self">self</span>)
-> <span class="kw-2">&</span><span class="kw">dyn </span>Any {
+<a href=#99 id=99 data-nosnippet>99</a> <span class="self">self
+<a href=#100 id=100 data-nosnippet>100</a> </span>}
+<a href=#101 id=101 data-nosnippet>101</a>
+<a href=#102 id=102 data-nosnippet>102</a> <span class="kw">fn
</span>table_names(<span class="kw-2">&</span><span
class="self">self</span>) -> Vec<String> {
+<a href=#103 id=103 data-nosnippet>103</a> <span
class="self">self</span>.tables
+<a href=#104 id=104 data-nosnippet>104</a> .iter()
+<a href=#105 id=105 data-nosnippet>105</a> .flat_map(|entry| {
+<a href=#106 id=106 data-nosnippet>106</a> <span class="kw">let
</span>table_name = entry.key().clone();
+<a href=#107 id=107 data-nosnippet>107</a> [table_name.clone()]
+<a href=#108 id=108 data-nosnippet>108</a> .into_iter()
+<a href=#109 id=109 data-nosnippet>109</a> .chain(
+<a href=#110 id=110 data-nosnippet>110</a>
MetadataTableType::all_types().map(<span class="kw">move
</span>|metadata_table_name| {
+<a href=#111 id=111 data-nosnippet>111</a> <span
class="macro">format!</span>(<span class="string">"{}${}"</span>, table_name,
metadata_table_name.as_str())
+<a href=#112 id=112 data-nosnippet>112</a> }),
+<a href=#113 id=113 data-nosnippet>113</a> )
+<a href=#114 id=114 data-nosnippet>114</a> })
+<a href=#115 id=115 data-nosnippet>115</a> .collect()
+<a href=#116 id=116 data-nosnippet>116</a> }
+<a href=#117 id=117 data-nosnippet>117</a>
+<a href=#118 id=118 data-nosnippet>118</a> <span class="kw">fn
</span>table_exist(<span class="kw-2">&</span><span
class="self">self</span>, name: <span class="kw-2">&</span>str) -> bool {
+<a href=#119 id=119 data-nosnippet>119</a> <span class="kw">if let
</span><span class="prelude-val">Some</span>((table_name, metadata_table_name))
= name.split_once(<span class="string">'$'</span>) {
+<a href=#120 id=120 data-nosnippet>120</a> <span
class="self">self</span>.tables.contains_key(table_name)
+<a href=#121 id=121 data-nosnippet>121</a> &&
MetadataTableType::try_from(metadata_table_name).is_ok()
+<a href=#122 id=122 data-nosnippet>122</a> } <span class="kw">else
</span>{
+<a href=#123 id=123 data-nosnippet>123</a> <span
class="self">self</span>.tables.contains_key(name)
+<a href=#124 id=124 data-nosnippet>124</a> }
+<a href=#125 id=125 data-nosnippet>125</a> }
<a href=#126 id=126 data-nosnippet>126</a>
-<a href=#127 id=127 data-nosnippet>127</a> <span
class="prelude-val">Ok</span>(<span class="self">self
-<a href=#128 id=128 data-nosnippet>128</a> </span>.tables
-<a href=#129 id=129 data-nosnippet>129</a> .get(name)
-<a href=#130 id=130 data-nosnippet>130</a> .cloned()
-<a href=#131 id=131 data-nosnippet>131</a> .map(|t| t <span
class="kw">as </span>Arc<<span class="kw">dyn </span>TableProvider>))
-<a href=#132 id=132 data-nosnippet>132</a> }
-<a href=#133 id=133
data-nosnippet>133</a>}</code></pre></div></section></main></body></html>
\ No newline at end of file
+<a href=#127 id=127 data-nosnippet>127</a> <span class="kw">async fn
</span>table(<span class="kw-2">&</span><span class="self">self</span>,
name: <span class="kw-2">&</span>str) -> DFResult<<span
class="prelude-ty">Option</span><Arc<<span class="kw">dyn
</span>TableProvider>>> {
+<a href=#128 id=128 data-nosnippet>128</a> <span class="kw">if let
</span><span class="prelude-val">Some</span>((table_name, metadata_table_name))
= name.split_once(<span class="string">'$'</span>) {
+<a href=#129 id=129 data-nosnippet>129</a> <span class="kw">let
</span>metadata_table_type =
+<a href=#130 id=130 data-nosnippet>130</a>
MetadataTableType::try_from(metadata_table_name).map_err(DataFusionError::Plan)<span
class="question-mark">?</span>;
+<a href=#131 id=131 data-nosnippet>131</a> <span class="kw">if let
</span><span class="prelude-val">Some</span>(table) = <span
class="self">self</span>.tables.get(table_name) {
+<a href=#132 id=132 data-nosnippet>132</a> <span class="kw">let
</span>metadata_table = table
+<a href=#133 id=133 data-nosnippet>133</a>
.metadata_table(metadata_table_type)
+<a href=#134 id=134 data-nosnippet>134</a> .<span
class="kw">await
+<a href=#135 id=135 data-nosnippet>135</a>
</span>.map_err(to_datafusion_error)<span class="question-mark">?</span>;
+<a href=#136 id=136 data-nosnippet>136</a> <span
class="kw">return </span><span class="prelude-val">Ok</span>(<span
class="prelude-val">Some</span>(Arc::new(metadata_table)));
+<a href=#137 id=137 data-nosnippet>137</a> } <span class="kw">else
</span>{
+<a href=#138 id=138 data-nosnippet>138</a> <span
class="kw">return </span><span class="prelude-val">Ok</span>(<span
class="prelude-val">None</span>);
+<a href=#139 id=139 data-nosnippet>139</a> }
+<a href=#140 id=140 data-nosnippet>140</a> }
+<a href=#141 id=141 data-nosnippet>141</a>
+<a href=#142 id=142 data-nosnippet>142</a> <span
class="prelude-val">Ok</span>(<span class="self">self
+<a href=#143 id=143 data-nosnippet>143</a> </span>.tables
+<a href=#144 id=144 data-nosnippet>144</a> .get(name)
+<a href=#145 id=145 data-nosnippet>145</a> .map(|entry|
entry.value().clone() <span class="kw">as </span>Arc<<span class="kw">dyn
</span>TableProvider>))
+<a href=#146 id=146 data-nosnippet>146</a> }
+<a href=#147 id=147 data-nosnippet>147</a>
+<a href=#148 id=148 data-nosnippet>148</a> <span class="kw">fn
</span>register_table(
+<a href=#149 id=149 data-nosnippet>149</a> <span
class="kw-2">&</span><span class="self">self</span>,
+<a href=#150 id=150 data-nosnippet>150</a> name: String,
+<a href=#151 id=151 data-nosnippet>151</a> table: Arc<<span
class="kw">dyn </span>TableProvider>,
+<a href=#152 id=152 data-nosnippet>152</a> ) -> DFResult<<span
class="prelude-ty">Option</span><Arc<<span class="kw">dyn
</span>TableProvider>>> {
+<a href=#153 id=153 data-nosnippet>153</a> <span class="comment">//
Check if table already exists
+<a href=#154 id=154 data-nosnippet>154</a> </span><span class="kw">if
</span><span class="self">self</span>.table_exist(name.as_str()) {
+<a href=#155 id=155 data-nosnippet>155</a> <span class="kw">return
</span><span class="prelude-val">Err</span>(DataFusionError::Execution(<span
class="macro">format!</span>(
+<a href=#156 id=156 data-nosnippet>156</a> <span
class="string">"Table {name} already exists"
+<a href=#157 id=157 data-nosnippet>157</a> </span>)));
+<a href=#158 id=158 data-nosnippet>158</a> }
+<a href=#159 id=159 data-nosnippet>159</a>
+<a href=#160 id=160 data-nosnippet>160</a> <span class="comment">//
Convert DataFusion schema to Iceberg schema
+<a href=#161 id=161 data-nosnippet>161</a> // DataFusion schemas don't
have field IDs, so we use the function that assigns them automatically
+<a href=#162 id=162 data-nosnippet>162</a> </span><span class="kw">let
</span>df_schema = table.schema();
+<a href=#163 id=163 data-nosnippet>163</a> <span class="kw">let
</span>iceberg_schema =
arrow_schema_to_schema_auto_assign_ids(df_schema.as_ref())
+<a href=#164 id=164 data-nosnippet>164</a>
.map_err(to_datafusion_error)<span class="question-mark">?</span>;
+<a href=#165 id=165 data-nosnippet>165</a>
+<a href=#166 id=166 data-nosnippet>166</a> <span class="comment">//
Create the table in the Iceberg catalog
+<a href=#167 id=167 data-nosnippet>167</a> </span><span class="kw">let
</span>table_creation = TableCreation::builder()
+<a href=#168 id=168 data-nosnippet>168</a> .name(name.clone())
+<a href=#169 id=169 data-nosnippet>169</a> .schema(iceberg_schema)
+<a href=#170 id=170 data-nosnippet>170</a> .build();
+<a href=#171 id=171 data-nosnippet>171</a>
+<a href=#172 id=172 data-nosnippet>172</a> <span class="kw">let
</span>catalog = <span class="self">self</span>.catalog.clone();
+<a href=#173 id=173 data-nosnippet>173</a> <span class="kw">let
</span>namespace = <span class="self">self</span>.namespace.clone();
+<a href=#174 id=174 data-nosnippet>174</a> <span class="kw">let
</span>tables = <span class="self">self</span>.tables.clone();
+<a href=#175 id=175 data-nosnippet>175</a> <span class="kw">let
</span>name_clone = name.clone();
+<a href=#176 id=176 data-nosnippet>176</a>
+<a href=#177 id=177 data-nosnippet>177</a> <span class="comment">// Use
tokio's spawn_blocking to handle the async work on a blocking thread pool
+<a href=#178 id=178 data-nosnippet>178</a> </span><span class="kw">let
</span>result = tokio::task::spawn_blocking(<span class="kw">move </span>|| {
+<a href=#179 id=179 data-nosnippet>179</a> <span class="comment">//
Create a new runtime handle to execute the async work
+<a href=#180 id=180 data-nosnippet>180</a> </span><span
class="kw">let </span>rt = tokio::runtime::Handle::current();
+<a href=#181 id=181 data-nosnippet>181</a> rt.block_on(<span
class="kw">async move </span>{
+<a href=#182 id=182 data-nosnippet>182</a> <span
class="comment">// Verify the input table is empty - CREATE TABLE only accepts
schema definition
+<a href=#183 id=183 data-nosnippet>183</a>
</span>ensure_table_is_empty(<span class="kw-2">&</span>table)
+<a href=#184 id=184 data-nosnippet>184</a> .<span
class="kw">await
+<a href=#185 id=185 data-nosnippet>185</a>
</span>.map_err(to_datafusion_error)<span class="question-mark">?</span>;
+<a href=#186 id=186 data-nosnippet>186</a>
+<a href=#187 id=187 data-nosnippet>187</a> catalog
+<a href=#188 id=188 data-nosnippet>188</a>
.create_table(<span class="kw-2">&</span>namespace, table_creation)
+<a href=#189 id=189 data-nosnippet>189</a> .<span
class="kw">await
+<a href=#190 id=190 data-nosnippet>190</a>
</span>.map_err(to_datafusion_error)<span class="question-mark">?</span>;
+<a href=#191 id=191 data-nosnippet>191</a>
+<a href=#192 id=192 data-nosnippet>192</a> <span
class="comment">// Create a new table provider using the catalog reference
+<a href=#193 id=193 data-nosnippet>193</a> </span><span
class="kw">let </span>table_provider = IcebergTableProvider::try_new(
+<a href=#194 id=194 data-nosnippet>194</a> catalog.clone(),
+<a href=#195 id=195 data-nosnippet>195</a>
namespace.clone(),
+<a href=#196 id=196 data-nosnippet>196</a>
name_clone.clone(),
+<a href=#197 id=197 data-nosnippet>197</a> )
+<a href=#198 id=198 data-nosnippet>198</a> .<span
class="kw">await
+<a href=#199 id=199 data-nosnippet>199</a>
</span>.map_err(to_datafusion_error)<span class="question-mark">?</span>;
+<a href=#200 id=200 data-nosnippet>200</a>
+<a href=#201 id=201 data-nosnippet>201</a> <span
class="comment">// Store the new table provider
+<a href=#202 id=202 data-nosnippet>202</a>
</span>tables.insert(name_clone, Arc::new(table_provider));
+<a href=#203 id=203 data-nosnippet>203</a>
+<a href=#204 id=204 data-nosnippet>204</a> <span
class="prelude-val">Ok</span>(<span class="prelude-val">None</span>)
+<a href=#205 id=205 data-nosnippet>205</a> })
+<a href=#206 id=206 data-nosnippet>206</a> });
+<a href=#207 id=207 data-nosnippet>207</a>
+<a href=#208 id=208 data-nosnippet>208</a> <span class="comment">//
Block on the spawned task to get the result
+<a href=#209 id=209 data-nosnippet>209</a> // This is safe because
spawn_blocking moves the blocking to a dedicated thread pool
+<a href=#210 id=210 data-nosnippet>210</a>
</span>futures::executor::block_on(result).map_err(|e| {
+<a href=#211 id=211 data-nosnippet>211</a>
DataFusionError::Execution(<span class="macro">format!</span>(<span
class="string">"Failed to create Iceberg table: {e}"</span>))
+<a href=#212 id=212 data-nosnippet>212</a> })<span
class="question-mark">?
+<a href=#213 id=213 data-nosnippet>213</a> </span>}
+<a href=#214 id=214 data-nosnippet>214</a>}
+<a href=#215 id=215 data-nosnippet>215</a>
+<a href=#216 id=216 data-nosnippet>216</a><span class="doccomment">///
Verifies that a table provider contains no data by scanning with LIMIT 1.
+<a href=#217 id=217 data-nosnippet>217</a>/// Returns an error if the table
has any rows.
+<a href=#218 id=218 data-nosnippet>218</a></span><span class="kw">async fn
</span>ensure_table_is_empty(table: <span class="kw-2">&</span>Arc<<span
class="kw">dyn </span>TableProvider>) -> <span
class="prelude-ty">Result</span><()> {
+<a href=#219 id=219 data-nosnippet>219</a> <span class="kw">let
</span>session_ctx = SessionContext::new();
+<a href=#220 id=220 data-nosnippet>220</a> <span class="kw">let
</span>exec_plan = table
+<a href=#221 id=221 data-nosnippet>221</a> .scan(<span
class="kw-2">&</span>session_ctx.state(), <span
class="prelude-val">None</span>, <span class="kw-2">&</span>[], <span
class="prelude-val">Some</span>(<span class="number">1</span>))
+<a href=#222 id=222 data-nosnippet>222</a> .<span class="kw">await
+<a href=#223 id=223 data-nosnippet>223</a> </span>.map_err(|e|
Error::new(ErrorKind::Unexpected, <span class="macro">format!</span>(<span
class="string">"Failed to scan table: {e}"</span>)))<span
class="question-mark">?</span>;
+<a href=#224 id=224 data-nosnippet>224</a>
+<a href=#225 id=225 data-nosnippet>225</a> <span class="kw">let
</span>task_ctx = Arc::new(TaskContext::default());
+<a href=#226 id=226 data-nosnippet>226</a> <span class="kw">let
</span>stream = exec_plan.execute(<span class="number">0</span>,
task_ctx).map_err(|e| {
+<a href=#227 id=227 data-nosnippet>227</a> Error::new(
+<a href=#228 id=228 data-nosnippet>228</a> ErrorKind::Unexpected,
+<a href=#229 id=229 data-nosnippet>229</a> <span
class="macro">format!</span>(<span class="string">"Failed to execute scan:
{e}"</span>),
+<a href=#230 id=230 data-nosnippet>230</a> )
+<a href=#231 id=231 data-nosnippet>231</a> })<span
class="question-mark">?</span>;
+<a href=#232 id=232 data-nosnippet>232</a>
+<a href=#233 id=233 data-nosnippet>233</a> <span class="kw">let
</span>batches: Vec<<span class="kw">_</span>> = stream.collect().<span
class="kw">await</span>;
+<a href=#234 id=234 data-nosnippet>234</a> <span class="kw">let
</span>has_data = batches
+<a href=#235 id=235 data-nosnippet>235</a> .into_iter()
+<a href=#236 id=236 data-nosnippet>236</a> .filter_map(|r| r.ok())
+<a href=#237 id=237 data-nosnippet>237</a> .any(|batch|
batch.num_rows() > <span class="number">0</span>);
+<a href=#238 id=238 data-nosnippet>238</a>
+<a href=#239 id=239 data-nosnippet>239</a> <span class="kw">if
</span>has_data {
+<a href=#240 id=240 data-nosnippet>240</a> <span class="kw">return
</span><span class="prelude-val">Err</span>(Error::new(
+<a href=#241 id=241 data-nosnippet>241</a> ErrorKind::Unexpected,
+<a href=#242 id=242 data-nosnippet>242</a> <span
class="string">"register_table does not support tables with data."</span>,
+<a href=#243 id=243 data-nosnippet>243</a> ));
+<a href=#244 id=244 data-nosnippet>244</a> }
+<a href=#245 id=245 data-nosnippet>245</a>
+<a href=#246 id=246 data-nosnippet>246</a> <span
class="prelude-val">Ok</span>(())
+<a href=#247 id=247 data-nosnippet>247</a>}
+<a href=#248 id=248 data-nosnippet>248</a>
+<a href=#249 id=249 data-nosnippet>249</a><span class="attr">#[cfg(test)]
+<a href=#250 id=250 data-nosnippet>250</a></span><span class="kw">mod
</span>tests {
+<a href=#251 id=251 data-nosnippet>251</a> <span class="kw">use
</span>std::collections::HashMap;
+<a href=#252 id=252 data-nosnippet>252</a> <span class="kw">use
</span>std::sync::Arc;
+<a href=#253 id=253 data-nosnippet>253</a>
+<a href=#254 id=254 data-nosnippet>254</a> <span class="kw">use
</span>datafusion::arrow::array::{Int32Array, StringArray};
+<a href=#255 id=255 data-nosnippet>255</a> <span class="kw">use
</span>datafusion::arrow::datatypes::{DataType, Field, Schema <span
class="kw">as </span>ArrowSchema};
+<a href=#256 id=256 data-nosnippet>256</a> <span class="kw">use
</span>datafusion::arrow::record_batch::RecordBatch;
+<a href=#257 id=257 data-nosnippet>257</a> <span class="kw">use
</span>datafusion::datasource::MemTable;
+<a href=#258 id=258 data-nosnippet>258</a> <span class="kw">use
</span>iceberg::memory::{MEMORY_CATALOG_WAREHOUSE, MemoryCatalogBuilder};
+<a href=#259 id=259 data-nosnippet>259</a> <span class="kw">use
</span>iceberg::{Catalog, CatalogBuilder, NamespaceIdent};
+<a href=#260 id=260 data-nosnippet>260</a> <span class="kw">use
</span>tempfile::TempDir;
+<a href=#261 id=261 data-nosnippet>261</a>
+<a href=#262 id=262 data-nosnippet>262</a> <span class="kw">use
super</span>::<span class="kw-2">*</span>;
+<a href=#263 id=263 data-nosnippet>263</a>
+<a href=#264 id=264 data-nosnippet>264</a> <span class="kw">async fn
</span>create_test_schema_provider() -> (IcebergSchemaProvider, TempDir) {
+<a href=#265 id=265 data-nosnippet>265</a> <span class="kw">let
</span>temp_dir = TempDir::new().unwrap();
+<a href=#266 id=266 data-nosnippet>266</a> <span class="kw">let
</span>warehouse_path = temp_dir.path().to_str().unwrap().to_string();
+<a href=#267 id=267 data-nosnippet>267</a>
+<a href=#268 id=268 data-nosnippet>268</a> <span class="kw">let
</span>catalog = MemoryCatalogBuilder::default()
+<a href=#269 id=269 data-nosnippet>269</a> .load(
+<a href=#270 id=270 data-nosnippet>270</a> <span
class="string">"memory"</span>,
+<a href=#271 id=271 data-nosnippet>271</a>
HashMap::from([(MEMORY_CATALOG_WAREHOUSE.to_string(), warehouse_path.clone())]),
+<a href=#272 id=272 data-nosnippet>272</a> )
+<a href=#273 id=273 data-nosnippet>273</a> .<span class="kw">await
+<a href=#274 id=274 data-nosnippet>274</a> </span>.unwrap();
+<a href=#275 id=275 data-nosnippet>275</a>
+<a href=#276 id=276 data-nosnippet>276</a> <span class="kw">let
</span>namespace = NamespaceIdent::new(<span
class="string">"test_ns"</span>.to_string());
+<a href=#277 id=277 data-nosnippet>277</a> catalog
+<a href=#278 id=278 data-nosnippet>278</a> .create_namespace(<span
class="kw-2">&</span>namespace, HashMap::new())
+<a href=#279 id=279 data-nosnippet>279</a> .<span class="kw">await
+<a href=#280 id=280 data-nosnippet>280</a> </span>.unwrap();
+<a href=#281 id=281 data-nosnippet>281</a>
+<a href=#282 id=282 data-nosnippet>282</a> <span class="kw">let
</span>provider = IcebergSchemaProvider::try_new(Arc::new(catalog), namespace)
+<a href=#283 id=283 data-nosnippet>283</a> .<span class="kw">await
+<a href=#284 id=284 data-nosnippet>284</a> </span>.unwrap();
+<a href=#285 id=285 data-nosnippet>285</a>
+<a href=#286 id=286 data-nosnippet>286</a> (provider, temp_dir)
+<a href=#287 id=287 data-nosnippet>287</a> }
+<a href=#288 id=288 data-nosnippet>288</a>
+<a href=#289 id=289 data-nosnippet>289</a> <span class="attr">#[tokio::test]
+<a href=#290 id=290 data-nosnippet>290</a> </span><span class="kw">async fn
</span>test_register_table_with_data_fails() {
+<a href=#291 id=291 data-nosnippet>291</a> <span class="kw">let
</span>(schema_provider, _temp_dir) = create_test_schema_provider().<span
class="kw">await</span>;
+<a href=#292 id=292 data-nosnippet>292</a>
+<a href=#293 id=293 data-nosnippet>293</a> <span class="comment">//
Create a MemTable with data
+<a href=#294 id=294 data-nosnippet>294</a> </span><span class="kw">let
</span>arrow_schema = Arc::new(ArrowSchema::new(<span class="macro">vec!</span>[
+<a href=#295 id=295 data-nosnippet>295</a> Field::new(<span
class="string">"id"</span>, DataType::Int32, <span
class="bool-val">false</span>),
+<a href=#296 id=296 data-nosnippet>296</a> Field::new(<span
class="string">"name"</span>, DataType::Utf8, <span
class="bool-val">true</span>),
+<a href=#297 id=297 data-nosnippet>297</a> ]));
+<a href=#298 id=298 data-nosnippet>298</a>
+<a href=#299 id=299 data-nosnippet>299</a> <span class="kw">let
</span>batch = RecordBatch::try_new(arrow_schema.clone(), <span
class="macro">vec!</span>[
+<a href=#300 id=300 data-nosnippet>300</a>
Arc::new(Int32Array::from(<span class="macro">vec!</span>[<span
class="number">1</span>, <span class="number">2</span>, <span
class="number">3</span>])),
+<a href=#301 id=301 data-nosnippet>301</a>
Arc::new(StringArray::from(<span class="macro">vec!</span>[<span
class="string">"Alice"</span>, <span class="string">"Bob"</span>, <span
class="string">"Charlie"</span>])),
+<a href=#302 id=302 data-nosnippet>302</a> ])
+<a href=#303 id=303 data-nosnippet>303</a> .unwrap();
+<a href=#304 id=304 data-nosnippet>304</a>
+<a href=#305 id=305 data-nosnippet>305</a> <span class="kw">let
</span>mem_table = MemTable::try_new(arrow_schema, <span
class="macro">vec!</span>[<span class="macro">vec!</span>[batch]]).unwrap();
+<a href=#306 id=306 data-nosnippet>306</a>
+<a href=#307 id=307 data-nosnippet>307</a> <span class="comment">//
Attempt to register the table with data - should fail
+<a href=#308 id=308 data-nosnippet>308</a> </span><span class="kw">let
</span>result = schema_provider.register_table(<span
class="string">"test_table"</span>.to_string(), Arc::new(mem_table));
+<a href=#309 id=309 data-nosnippet>309</a>
+<a href=#310 id=310 data-nosnippet>310</a> <span
class="macro">assert!</span>(result.is_err());
+<a href=#311 id=311 data-nosnippet>311</a> <span class="kw">let
</span>err = result.unwrap_err();
+<a href=#312 id=312 data-nosnippet>312</a> <span
class="macro">assert!</span>(
+<a href=#313 id=313 data-nosnippet>313</a> err.to_string()
+<a href=#314 id=314 data-nosnippet>314</a> .contains(<span
class="string">"register_table does not support tables with data."</span>),
+<a href=#315 id=315 data-nosnippet>315</a> <span
class="string">"Expected error about tables with data, got: {err}"</span>,
+<a href=#316 id=316 data-nosnippet>316</a> );
+<a href=#317 id=317 data-nosnippet>317</a> }
+<a href=#318 id=318 data-nosnippet>318</a>
+<a href=#319 id=319 data-nosnippet>319</a> <span class="attr">#[tokio::test]
+<a href=#320 id=320 data-nosnippet>320</a> </span><span class="kw">async fn
</span>test_register_empty_table_succeeds() {
+<a href=#321 id=321 data-nosnippet>321</a> <span class="kw">let
</span>(schema_provider, _temp_dir) = create_test_schema_provider().<span
class="kw">await</span>;
+<a href=#322 id=322 data-nosnippet>322</a>
+<a href=#323 id=323 data-nosnippet>323</a> <span class="comment">//
Create an empty MemTable (schema only, no data rows)
+<a href=#324 id=324 data-nosnippet>324</a> </span><span class="kw">let
</span>arrow_schema = Arc::new(ArrowSchema::new(<span class="macro">vec!</span>[
+<a href=#325 id=325 data-nosnippet>325</a> Field::new(<span
class="string">"id"</span>, DataType::Int32, <span
class="bool-val">false</span>),
+<a href=#326 id=326 data-nosnippet>326</a> Field::new(<span
class="string">"name"</span>, DataType::Utf8, <span
class="bool-val">true</span>),
+<a href=#327 id=327 data-nosnippet>327</a> ]));
+<a href=#328 id=328 data-nosnippet>328</a>
+<a href=#329 id=329 data-nosnippet>329</a> <span class="comment">//
Create an empty batch (0 rows) - MemTable requires at least one partition
+<a href=#330 id=330 data-nosnippet>330</a> </span><span class="kw">let
</span>empty_batch = RecordBatch::new_empty(arrow_schema.clone());
+<a href=#331 id=331 data-nosnippet>331</a> <span class="kw">let
</span>mem_table = MemTable::try_new(arrow_schema, <span
class="macro">vec!</span>[<span
class="macro">vec!</span>[empty_batch]]).unwrap();
+<a href=#332 id=332 data-nosnippet>332</a>
+<a href=#333 id=333 data-nosnippet>333</a> <span class="comment">//
Attempt to register the empty table - should succeed
+<a href=#334 id=334 data-nosnippet>334</a> </span><span class="kw">let
</span>result = schema_provider.register_table(<span
class="string">"empty_table"</span>.to_string(), Arc::new(mem_table));
+<a href=#335 id=335 data-nosnippet>335</a>
+<a href=#336 id=336 data-nosnippet>336</a> <span
class="macro">assert!</span>(result.is_ok(), <span class="string">"Expected
success, got: {result:?}"</span>);
+<a href=#337 id=337 data-nosnippet>337</a>
+<a href=#338 id=338 data-nosnippet>338</a> <span class="comment">//
Verify the table was registered
+<a href=#339 id=339 data-nosnippet>339</a> </span><span
class="macro">assert!</span>(schema_provider.table_exist(<span
class="string">"empty_table"</span>));
+<a href=#340 id=340 data-nosnippet>340</a> }
+<a href=#341 id=341 data-nosnippet>341</a>
+<a href=#342 id=342 data-nosnippet>342</a> <span class="attr">#[tokio::test]
+<a href=#343 id=343 data-nosnippet>343</a> </span><span class="kw">async fn
</span>test_register_duplicate_table_fails() {
+<a href=#344 id=344 data-nosnippet>344</a> <span class="kw">let
</span>(schema_provider, _temp_dir) = create_test_schema_provider().<span
class="kw">await</span>;
+<a href=#345 id=345 data-nosnippet>345</a>
+<a href=#346 id=346 data-nosnippet>346</a> <span class="comment">//
Create empty MemTables
+<a href=#347 id=347 data-nosnippet>347</a> </span><span class="kw">let
</span>arrow_schema = Arc::new(ArrowSchema::new(<span
class="macro">vec!</span>[Field::new(
+<a href=#348 id=348 data-nosnippet>348</a> <span
class="string">"id"</span>,
+<a href=#349 id=349 data-nosnippet>349</a> DataType::Int32,
+<a href=#350 id=350 data-nosnippet>350</a> <span
class="bool-val">false</span>,
+<a href=#351 id=351 data-nosnippet>351</a> )]));
+<a href=#352 id=352 data-nosnippet>352</a>
+<a href=#353 id=353 data-nosnippet>353</a> <span class="kw">let
</span>empty_batch1 = RecordBatch::new_empty(arrow_schema.clone());
+<a href=#354 id=354 data-nosnippet>354</a> <span class="kw">let
</span>empty_batch2 = RecordBatch::new_empty(arrow_schema.clone());
+<a href=#355 id=355 data-nosnippet>355</a> <span class="kw">let
</span>mem_table1 = MemTable::try_new(arrow_schema.clone(), <span
class="macro">vec!</span>[<span
class="macro">vec!</span>[empty_batch1]]).unwrap();
+<a href=#356 id=356 data-nosnippet>356</a> <span class="kw">let
</span>mem_table2 = MemTable::try_new(arrow_schema, <span
class="macro">vec!</span>[<span
class="macro">vec!</span>[empty_batch2]]).unwrap();
+<a href=#357 id=357 data-nosnippet>357</a>
+<a href=#358 id=358 data-nosnippet>358</a> <span class="comment">//
Register first table - should succeed
+<a href=#359 id=359 data-nosnippet>359</a> </span><span class="kw">let
</span>result1 = schema_provider.register_table(<span
class="string">"dup_table"</span>.to_string(), Arc::new(mem_table1));
+<a href=#360 id=360 data-nosnippet>360</a> <span
class="macro">assert!</span>(result1.is_ok());
+<a href=#361 id=361 data-nosnippet>361</a>
+<a href=#362 id=362 data-nosnippet>362</a> <span class="comment">//
Register second table with same name - should fail
+<a href=#363 id=363 data-nosnippet>363</a> </span><span class="kw">let
</span>result2 = schema_provider.register_table(<span
class="string">"dup_table"</span>.to_string(), Arc::new(mem_table2));
+<a href=#364 id=364 data-nosnippet>364</a> <span
class="macro">assert!</span>(result2.is_err());
+<a href=#365 id=365 data-nosnippet>365</a> <span class="kw">let
</span>err = result2.unwrap_err();
+<a href=#366 id=366 data-nosnippet>366</a> <span
class="macro">assert!</span>(
+<a href=#367 id=367 data-nosnippet>367</a>
err.to_string().contains(<span class="string">"already exists"</span>),
+<a href=#368 id=368 data-nosnippet>368</a> <span
class="string">"Expected error about table already existing, got: {err}"</span>,
+<a href=#369 id=369 data-nosnippet>369</a> );
+<a href=#370 id=370 data-nosnippet>370</a> }
+<a href=#371 id=371
data-nosnippet>371</a>}</code></pre></div></section></main></body></html>
\ No newline at end of file
diff --git a/api/src/iceberg_sqllogictest/engine/datafusion.rs.html
b/api/src/iceberg_sqllogictest/engine/datafusion.rs.html
index 6992a74eb..9bedeeaf4 100644
--- a/api/src/iceberg_sqllogictest/engine/datafusion.rs.html
+++ b/api/src/iceberg_sqllogictest/engine/datafusion.rs.html
@@ -93,72 +93,45 @@
<a href=#93 id=93 data-nosnippet>93</a> </span><span class="kw">let
</span>namespace = NamespaceIdent::new(<span
class="string">"default"</span>.to_string());
<a href=#94 id=94 data-nosnippet>94</a> catalog.create_namespace(<span
class="kw-2">&</span>namespace, HashMap::new()).<span
class="kw">await</span><span class="question-mark">?</span>;
<a href=#95 id=95 data-nosnippet>95</a>
-<a href=#96 id=96 data-nosnippet>96</a> <span class="comment">// Create
test tables
-<a href=#97 id=97 data-nosnippet>97</a> </span><span
class="self">Self</span>::create_unpartitioned_table(<span
class="kw-2">&</span>catalog, <span
class="kw-2">&</span>namespace).<span class="kw">await</span><span
class="question-mark">?</span>;
-<a href=#98 id=98 data-nosnippet>98</a> <span
class="self">Self</span>::create_partitioned_table(<span
class="kw-2">&</span>catalog, <span
class="kw-2">&</span>namespace).<span class="kw">await</span><span
class="question-mark">?</span>;
-<a href=#99 id=99 data-nosnippet>99</a>
-<a href=#100 id=100 data-nosnippet>100</a> <span
class="prelude-val">Ok</span>(Arc::new(
-<a href=#101 id=101 data-nosnippet>101</a>
IcebergCatalogProvider::try_new(Arc::new(catalog)).<span
class="kw">await</span><span class="question-mark">?</span>,
-<a href=#102 id=102 data-nosnippet>102</a> ))
-<a href=#103 id=103 data-nosnippet>103</a> }
-<a href=#104 id=104 data-nosnippet>104</a>
-<a href=#105 id=105 data-nosnippet>105</a> <span class="doccomment">///
Create an unpartitioned test table with id and name columns
-<a href=#106 id=106 data-nosnippet>106</a> /// TODO: this can be removed
when we support CREATE TABLE
-<a href=#107 id=107 data-nosnippet>107</a> </span><span class="kw">async fn
</span>create_unpartitioned_table(
+<a href=#96 id=96 data-nosnippet>96</a> <span class="comment">// Create
partitioned test table (unpartitioned tables are now created via SQL)
+<a href=#97 id=97 data-nosnippet>97</a> </span><span
class="self">Self</span>::create_partitioned_table(<span
class="kw-2">&</span>catalog, <span
class="kw-2">&</span>namespace).<span class="kw">await</span><span
class="question-mark">?</span>;
+<a href=#98 id=98 data-nosnippet>98</a>
+<a href=#99 id=99 data-nosnippet>99</a> <span
class="prelude-val">Ok</span>(Arc::new(
+<a href=#100 id=100 data-nosnippet>100</a>
IcebergCatalogProvider::try_new(Arc::new(catalog)).<span
class="kw">await</span><span class="question-mark">?</span>,
+<a href=#101 id=101 data-nosnippet>101</a> ))
+<a href=#102 id=102 data-nosnippet>102</a> }
+<a href=#103 id=103 data-nosnippet>103</a>
+<a href=#104 id=104 data-nosnippet>104</a> <span class="doccomment">///
Create a partitioned test table with id, category, and value columns
+<a href=#105 id=105 data-nosnippet>105</a> /// Partitioned by category
using identity transform
+<a href=#106 id=106 data-nosnippet>106</a> /// TODO: this can be removed
when we support CREATE EXTERNAL TABLE
+<a href=#107 id=107 data-nosnippet>107</a> </span><span class="kw">async fn
</span>create_partitioned_table(
<a href=#108 id=108 data-nosnippet>108</a> catalog: <span
class="kw-2">&</span><span class="kw">impl </span>Catalog,
<a href=#109 id=109 data-nosnippet>109</a> namespace: <span
class="kw-2">&</span>NamespaceIdent,
<a href=#110 id=110 data-nosnippet>110</a> ) -> anyhow::Result<()>
{
<a href=#111 id=111 data-nosnippet>111</a> <span class="kw">let
</span>schema = Schema::builder()
<a href=#112 id=112 data-nosnippet>112</a> .with_fields(<span
class="macro">vec!</span>[
<a href=#113 id=113 data-nosnippet>113</a>
NestedField::required(<span class="number">1</span>, <span
class="string">"id"</span>, Type::Primitive(PrimitiveType::Int)).into(),
-<a href=#114 id=114 data-nosnippet>114</a>
NestedField::optional(<span class="number">2</span>, <span
class="string">"name"</span>, Type::Primitive(PrimitiveType::String)).into(),
-<a href=#115 id=115 data-nosnippet>115</a> ])
-<a href=#116 id=116 data-nosnippet>116</a> .build()<span
class="question-mark">?</span>;
-<a href=#117 id=117 data-nosnippet>117</a>
-<a href=#118 id=118 data-nosnippet>118</a> catalog
-<a href=#119 id=119 data-nosnippet>119</a> .create_table(
-<a href=#120 id=120 data-nosnippet>120</a> namespace,
-<a href=#121 id=121 data-nosnippet>121</a>
TableCreation::builder()
-<a href=#122 id=122 data-nosnippet>122</a> .name(<span
class="string">"test_unpartitioned_table"</span>.to_string())
-<a href=#123 id=123 data-nosnippet>123</a> .schema(schema)
-<a href=#124 id=124 data-nosnippet>124</a> .build(),
-<a href=#125 id=125 data-nosnippet>125</a> )
-<a href=#126 id=126 data-nosnippet>126</a> .<span
class="kw">await</span><span class="question-mark">?</span>;
-<a href=#127 id=127 data-nosnippet>127</a>
-<a href=#128 id=128 data-nosnippet>128</a> <span
class="prelude-val">Ok</span>(())
-<a href=#129 id=129 data-nosnippet>129</a> }
-<a href=#130 id=130 data-nosnippet>130</a>
-<a href=#131 id=131 data-nosnippet>131</a> <span class="doccomment">///
Create a partitioned test table with id, category, and value columns
-<a href=#132 id=132 data-nosnippet>132</a> /// Partitioned by category
using identity transform
-<a href=#133 id=133 data-nosnippet>133</a> /// TODO: this can be removed
when we support CREATE TABLE
-<a href=#134 id=134 data-nosnippet>134</a> </span><span class="kw">async fn
</span>create_partitioned_table(
-<a href=#135 id=135 data-nosnippet>135</a> catalog: <span
class="kw-2">&</span><span class="kw">impl </span>Catalog,
-<a href=#136 id=136 data-nosnippet>136</a> namespace: <span
class="kw-2">&</span>NamespaceIdent,
-<a href=#137 id=137 data-nosnippet>137</a> ) -> anyhow::Result<()>
{
-<a href=#138 id=138 data-nosnippet>138</a> <span class="kw">let
</span>schema = Schema::builder()
-<a href=#139 id=139 data-nosnippet>139</a> .with_fields(<span
class="macro">vec!</span>[
-<a href=#140 id=140 data-nosnippet>140</a>
NestedField::required(<span class="number">1</span>, <span
class="string">"id"</span>, Type::Primitive(PrimitiveType::Int)).into(),
-<a href=#141 id=141 data-nosnippet>141</a>
NestedField::required(<span class="number">2</span>, <span
class="string">"category"</span>,
Type::Primitive(PrimitiveType::String)).into(),
-<a href=#142 id=142 data-nosnippet>142</a>
NestedField::optional(<span class="number">3</span>, <span
class="string">"value"</span>, Type::Primitive(PrimitiveType::String)).into(),
-<a href=#143 id=143 data-nosnippet>143</a> ])
-<a href=#144 id=144 data-nosnippet>144</a> .build()<span
class="question-mark">?</span>;
-<a href=#145 id=145 data-nosnippet>145</a>
-<a href=#146 id=146 data-nosnippet>146</a> <span class="kw">let
</span>partition_spec = UnboundPartitionSpec::builder()
-<a href=#147 id=147 data-nosnippet>147</a> .with_spec_id(<span
class="number">0</span>)
-<a href=#148 id=148 data-nosnippet>148</a>
.add_partition_field(<span class="number">2</span>, <span
class="string">"category"</span>, Transform::Identity)<span
class="question-mark">?
-<a href=#149 id=149 data-nosnippet>149</a> </span>.build();
-<a href=#150 id=150 data-nosnippet>150</a>
-<a href=#151 id=151 data-nosnippet>151</a> catalog
-<a href=#152 id=152 data-nosnippet>152</a> .create_table(
-<a href=#153 id=153 data-nosnippet>153</a> namespace,
-<a href=#154 id=154 data-nosnippet>154</a>
TableCreation::builder()
-<a href=#155 id=155 data-nosnippet>155</a> .name(<span
class="string">"test_partitioned_table"</span>.to_string())
-<a href=#156 id=156 data-nosnippet>156</a> .schema(schema)
-<a href=#157 id=157 data-nosnippet>157</a>
.partition_spec(partition_spec)
-<a href=#158 id=158 data-nosnippet>158</a> .build(),
-<a href=#159 id=159 data-nosnippet>159</a> )
-<a href=#160 id=160 data-nosnippet>160</a> .<span
class="kw">await</span><span class="question-mark">?</span>;
-<a href=#161 id=161 data-nosnippet>161</a>
-<a href=#162 id=162 data-nosnippet>162</a> <span
class="prelude-val">Ok</span>(())
-<a href=#163 id=163 data-nosnippet>163</a> }
-<a href=#164 id=164
data-nosnippet>164</a>}</code></pre></div></section></main></body></html>
\ No newline at end of file
+<a href=#114 id=114 data-nosnippet>114</a>
NestedField::required(<span class="number">2</span>, <span
class="string">"category"</span>,
Type::Primitive(PrimitiveType::String)).into(),
+<a href=#115 id=115 data-nosnippet>115</a>
NestedField::optional(<span class="number">3</span>, <span
class="string">"value"</span>, Type::Primitive(PrimitiveType::String)).into(),
+<a href=#116 id=116 data-nosnippet>116</a> ])
+<a href=#117 id=117 data-nosnippet>117</a> .build()<span
class="question-mark">?</span>;
+<a href=#118 id=118 data-nosnippet>118</a>
+<a href=#119 id=119 data-nosnippet>119</a> <span class="kw">let
</span>partition_spec = UnboundPartitionSpec::builder()
+<a href=#120 id=120 data-nosnippet>120</a> .with_spec_id(<span
class="number">0</span>)
+<a href=#121 id=121 data-nosnippet>121</a>
.add_partition_field(<span class="number">2</span>, <span
class="string">"category"</span>, Transform::Identity)<span
class="question-mark">?
+<a href=#122 id=122 data-nosnippet>122</a> </span>.build();
+<a href=#123 id=123 data-nosnippet>123</a>
+<a href=#124 id=124 data-nosnippet>124</a> catalog
+<a href=#125 id=125 data-nosnippet>125</a> .create_table(
+<a href=#126 id=126 data-nosnippet>126</a> namespace,
+<a href=#127 id=127 data-nosnippet>127</a>
TableCreation::builder()
+<a href=#128 id=128 data-nosnippet>128</a> .name(<span
class="string">"test_partitioned_table"</span>.to_string())
+<a href=#129 id=129 data-nosnippet>129</a> .schema(schema)
+<a href=#130 id=130 data-nosnippet>130</a>
.partition_spec(partition_spec)
+<a href=#131 id=131 data-nosnippet>131</a> .build(),
+<a href=#132 id=132 data-nosnippet>132</a> )
+<a href=#133 id=133 data-nosnippet>133</a> .<span
class="kw">await</span><span class="question-mark">?</span>;
+<a href=#134 id=134 data-nosnippet>134</a>
+<a href=#135 id=135 data-nosnippet>135</a> <span
class="prelude-val">Ok</span>(())
+<a href=#136 id=136 data-nosnippet>136</a> }
+<a href=#137 id=137
data-nosnippet>137</a>}</code></pre></div></section></main></body></html>
\ No newline at end of file
